|
Libey on RFM value Donald R. Libey
Table of Contents
Introduction 1
RECENCY 5 Real-Life Recency Characteristics 5 Recency and Intuitive Knowledge 10 Recency and Time 11 Recency and the Customer Database 14 Recency Segmentation 18 Recency Velocity 25 Recency Momentum 27 Recency Reconciliation 30 Recency and List Rentals 34 Recency and Inquiries 40 Recency and Customer Attrition Rates 42 Recency and Attrition Recovery 46 Recency and Multiple Customers 47 Recency and Why It Isn’t Enough 49
FREQUENCY 53 Real-Life Frequency Characteristics 53 Frequency and Product Usage 58 Frequency and Product Pricing 61 Frequency and Customer Service Frequency and the Marketing Hierarchy 70 Frequency and Product Propensity 71 Frequency and Price Changes 73 Frequency and Quantity Price Breaks 73 Frequency and Advertising Space 75 Frequency Velocity 76 Frequency Momentum 79 Frequency and Acquisitions 80 Frequency Reconciliation 82 Frequency Segmentation 85 Frequency and List Rentals 88 Frequency and Offer Life and Half-Life 89 Frequency and the Lapsing Customer 91 Frequency and Why It’s Not Quite Enough
MONETARY VALUE 98 Real-Life Monetary Value Characteristics Monetary Value and Segmentation 101 Monetary Value Velocity 105 Monetary Value Momentum 107 Monetary Value Reconciliation 109 Monetary Value and List Rentals 111 Monetary Value and Advertising Frequency Monetary Value and Customer Attrition Monetary Value and Product Propensity Monetary Value and Pricing 120 Monetary Value and Advertising 123 Monetary Value and Why It Isn’t Enough
OPTIMAL RFM 127 Optimizing Recency, Frequency and Monetary Value 127 Optimized RFM Plus SIC Codes 128 Optimized RFM Plus Geography 129 Optimized RFM Plus Financial Information Optimized RFM Plus Size 132 Optimized RFM Plus List Source 133 Optimized RFM Plus Age 135 Optimal Analyses 136
Copyright 138
Index 139
Direct marketing is a formulaic business. It is consistently predictable, and financial results can be projected with a high degree of accuracy and reliability. Knowing the formulas is the hallmark of the experienced practitioner of this largest of all advertising disciplines. Applying the formulas along with adequate levels of capital to create above average margin is the hallmark of the master entrepreneur of direct marketing. You can have all the experience and all the money, but if the basic formulas are not known and understood, then margin will elude you as sure as night follows day. Many formulas exist in direct marketing. They change based on numerous variables and externalities; however, only the percentages change, only the relationships. The relativities and the unities are preserved. Understanding the relativities and the unities is the key to creating margin in this complex business. This is a book about the formulaic nature of direct marketing. It is a book about the cardinal formulas: recency, frequency and monetary value. And it is a book about the infinite relationships among those three foundation elements. This book was written as a response to a jarring recognition. Few direct marketers, either consumer or business to business, thoroughly understand or expertly apply recency, frequency and monetary value in actual practice. In recent years, thousands of direct marketing conference and seminar participants have admitted having little or no familiarity or understanding of these basic foundation blocks of direct marketing. Once the primary tools of the apprentice direct marketer, these bedrock formulas have fallen into increasing obscurity. No direct marketer can be fully furnished without a facile understanding of recency, frequency and monetary value. To conceive such a void is analogous to a physician having no understanding of a patient’s temperature, pulse, or respiration. Generations of financially healthy direct marketers have retired to the comfort of Boca Raton having mastered recency, frequency and monetary value early in their careers. Surely, in the hostile and unhealthy economic environment of the future of direct marketing, few practitioners will follow the past-masters to wealth and success without first achieving their own healthy mastery of these three cardinal vital signs. The advances in direct marketing have been so great and the growth has been so phenomenal that thousands of corporations have entered the field. Without the learning curve of the basics, these organizations have often thrown enormous amounts of money into direct marketing without an adequate understanding of the measurements and the formulaic expectations. Consequently, luck dominates; either they get lucky or they go broke. At the same time, technological advancements, particularly in the areas of management and analytical software and in database technology, have embedded the basic functions of recency, frequency and monetary value to the point where new practitioners are unaware of their existence. Unless the three elements are unearthed and exposed to the light of understanding, they run the risk of being permanently buried within programming and forever out of the bright light of knowledge and wisdom. But even among long-established, experienced direct marketers, the ignorance of recency, frequency and monetary value is surprisingly high. At a time of industry maturity where one would expect a near 100% use of recency, frequency and monetary value by established direct marketing companies, the reality may, in fact, be less than 25%. The greatest number use recency; a smaller number use combined recency and frequency; the smallest number use full recency, frequency and monetary value. Our estimate, based wholly on unscientific, informal, personal surveys, leads us to believe that possibly only 10% of established U.S. direct marketing companies are practicing full recency, frequency and monetary value analyses. That would be very surprising in a nation renowned world-wide for its sophistication, leadership and technological dominance of the direct marketing discipline. In an ideal world, recency, frequency and monetary value would flow from the point of order entry in real time; analytic reports would be available to marketers at any moment in time, providing a snapshot of the dynamics of customer response in a continuing, replicable flow of updated, accurate information distilled to useful knowledge. Unfortunately, the ideal world seldom is found. Many direct marketers are only able to complete studies of recency, frequency and monetary value in isolation; the analyses cannot be replicated without starting over from scratch. A one-time picture is developed, but no capacity exists to take updated pictures. Nothing is less useful than knowing what happened six months ago and not knowing what is happening today. To be useful, recency, frequency and monetary value must evolve hour by hour, day by day, week by week, month by month, quarter by quarter, and year by year. The value lies in the ability to see change over time; indeed, that may be the only value as it becomes a replicable measurement of consistently improving profitability created from increasingly better decisions.
Recency is when a customer last purchased. It is an indicator of timing and freshness. As an aggregate of all customers, it is a primary measurement of business vitality. By itself, it is arguably the most important indicator of the cardinal three — recency, frequency and monetary value (RFM). Recency describes a state of a customer’s being. If you think about it, there are really only three states of being to describe customers. All people who have never bought from you are future customers; their state is potential. All people who have bought from you are current customers; their state is active. All people who bought in the past but are no longer buying are old customers; their state is inactive. The process of direct marketing uses the vast universe of potential customers to create a growing universe of active customers. Keeping those active customers active is the constant objective; otherwise, they dissipate into the inactive state. Only two sources exist from which to create active customers. They are recruited from the universe of potential customers (people who have never bought from you) or they are enticed back from the inactive, old customer state (people who bought from you one or more times but have stopped buying for whatever reason). Logic would dictate that the true definition of direct or any other branch of marketing is creating and keeping customers. Everything is aimed at filling the active state with as many customers as possible, old and new. If every customer would buy every day, you would have a perpetual motion money machine. Every customer would be recent and life would be simple. Alas, such is not the stuff of which reality is made. It is difficult to get people to buy from you for the first time, and it is hard to keep them buying from you ever after. Only the rare master marketer is able to squeeze thousands upon thousands of new and old customers into the active state. Recency also describes the state of the customer’s awareness. Differing levels and intensities of awareness exist in the mind of the customer at the same time, often involving hundreds of products and purchase relationships. A customer who routinely purchases The Wall Street Journal every morning exhibits a back-of-the-mind awareness that drives a regular state of recency. Reading The Wall Street Journal has become a habit. The same customer who purchases U.S. News & World Report now and then based on impulse exhibits a middle-of-the-mind awareness that drives a periodic state of recency. And the same customer who heard about a special feature article in this month’s issue of Forbes and rushes out to specifically buy a copy of Forbes exhibits a top-of-the-mind awareness that drives an episodic state of recency. The customer has a high recency level relative to The Wall Street Journal, a middle recency level relative to U.S. News & World Report and a low recency level relative to Forbes. However, the same customer has a low intensity relative to The Wall Street Journal, a middle intensity relative to U.S. News & World Report, and a high intensity relative to Forbes. Interestingly, the high intensity purchase of Forbes may only consist of one issue while the low intensity purchase of The Wall Street Journal may consist of 250 daily issues. This dichotomy raises the very interesting question, "Recency for what reason?" Recency that is episodic is generally the least valuable indicator of customer loyalty. Recency that is periodic is better, but not optimum. Recency that is regular is coveted by marketers. Another way to view recency is to consider it an indication of affirmation. If a customer purchases a pair of shoes from your shoe company once in a lifetime, that customer has made an episodic affirmation of loyalty. If, however, that customer purchases a pair of shoes from your company every ten years or so, that customer has made a periodic affirmation of loyalty. Best of all, if that customer purchases a pair of shoes from your company every year, that customer has made a regular affirmation of loyalty. You cannot count on the episodic loyalty; you can only marginally count on the periodic loyalty; to a much greater degree, you can count on the regular affirmation of loyalty. Marketers, then, are constantly searching for ways to create regular affirmations of loyalty. Master marketers take this process one level higher and constantly search for ways to create high intensity, regular affirmations of loyalty. A customer who buys The Wall Street Journal every day and exhibits a low intensity is affirming loyalty. Raised to the next level — high intensity — that customer will not only constantly reinforce self-loyalty but will bring additional customers into a state of regular recency. That is the coveted customer. In this manner of thinking about recency, it becomes clear that recency is, in a way, regeneration. Every time a purchase is made, the customer is regenerating. If the regeneration process is static, the level of intensity is low; if the regeneration process is dynamic, the level of intensity is high. A customer who buys The Wall Street Journal every day only for news may exhibit a low intensity, static regeneration, affirmation of loyalty. But a customer who buys because of a desire to track the price of stock that is owned exhibits a higher intensity, more dynamic regeneration, affirmation of loyalty. The desire for that customer to be recent has a reason. In other words, for that one customer, an answer to the question, "Recency for what reason?" has been found. Recency that is compelled is preferable to recency that is non-compelled. Among coffee drinkers, dependency upon caffeine compels regular recency of a higher intensity than that of non-coffee drinkers who purchase coffee only for periodic or episodic dinner guests. As a marketer of coffee, you can count on the daily, compelling need for caffeine to regenerate the regular affirmation of loyalty and recency of purchase. The recency ideal, then, is a compelling, dynamic, high intensity, regular affirmation of loyalty. Simply put, customers who are energized and compelled to demonstrate their loyalty through continual purchases are superb customers and the ones you want to create and keep. Recency and Intuitive Knowledge Modern marketing practices are the outgrowth of old-fashioned selling knowledge that the peddlers understood and used every day. Often denigrated by revisionist history, the peddler was, perhaps, the most efficient marketer in our commercial history. The customer database existed in memory and was updated on every sales call. The peddler knew who purchased what on the last trip, what was asked for on the last trip, and could surmise accurately what would likely be bought on the current trip. For each customer on the route, the peddler amassed large amounts of information, from the customers themselves and from listening to the gossip of neighbors. All of that knowledge was filed in memory and used to predict the next purchase of individual customers and to gauge the likely product interest and dollar value of the sale. The end result was that the peddler almost always made a sale at every stop and the recency of the customer was kept dynamic, energized and, more often than not, compelled through inside knowledge. As we enter the twenty-first century, it is no longer acceptable or politically correct to refer to ourselves as peddlers; yet, that is exactly what we are, albeit on a larger, technological route. Nothing has yet replaced intuitive knowledge of the customer as a primary characteristic of successful marketing. We are still traveling down a dirt road, riding a horse-drawn wagon, catching a glimpse of a farmhouse in the distance and saying to ourselves, "Let’s see . . . The last time we were here they bought two pots, an iron skillet, twelve spools of thread, six needles and a bottle of liver tonic." The demand for intuitive knowledge about individual customers has evolved from the peddler’s level of near-perfect, useful knowledge to a mass marketing, broad and generalized homogenization of pseudo-knowledge in the decades from 1950 to 1980 to today’s new century frenzy for individuated customer knowledge extending through layers of demographic and lifestyle data maintained in sophisticated, relational databases. We have, of course, only become better peddlers. At least, we think we are. Recency, for the peddler, meant that every customer had to buy on every trip. The wagon passed by only once or twice a year. Keeping each customer recent meant making a sale on every pass. Today, we have wagons that arrive in the mail as often as profitable; we have electronic wagons that millions of customers can browse through at will; we have interactive wagons that stop by the farm on demand. Recency today is more recent. Intuitive knowledge today is more recent. At what interval of time should a customer purchase from you in order to meet the ideal of a compelling, dynamic, high intensity, regular affirmation of loyalty? If you are The Wall Street Journal, recency time can be daily for newsstand buyers or annually for subscription buyers. Recency time is a function of the preferred method of purchasing and delivery of the product. For those customers who prefer to pick up the paper daily and pay cash, recency is daily; those who prefer to subscribe for a year and receive the paper in the mail, recency is annual. Viewed this way, recency is when money is paid. Right away, several interesting factors can be seen. If recency is daily, there is always the risk that the customer will buy another paper instead. To reduce that risk, the annual subscription locks in loyalty and eliminates the need for a daily reaffirmation of loyalty. Daily recency also requires that the customer has cash each day; if not, a sale is missed. Annual recency eliminates lost sales due to lack of pocket change; the paper arrives regardless of cash on hand. Perhaps annual recency is more stable than daily recency. But a newspaper is vastly different from other product purchases. It is, by definition, a daily recency product just as a weekly magazine is a weekly recency product and a monthly magazine is a monthly recency product. How do we equate recency and time to the thousands of customer relationships and products that enter into those relationships? When in time must a customer buy to be optimally recent? As we have seen, if the product is newspapers, the optimal recency time is daily. But what if the product is bridges across the Mississippi River? In the bridge-building business, a recent customer may be every 100 years. Recency and time are directly linked to product usage rate. If you have not had a newspaper purchase from a known customer in 30 days, that customer is probably reading some other newspaper. But if a state hasn’t bought a bridge from you in the last 50 years, you may be getting an order sometime in the next five decades. Thinking only of recency, what is an acceptable period of time for affirming loyalty through purchasing of the following products? peanuts automobiles IRAs surgery dogs flowers term life golf clubs checks termite control bread fax paper watches fly rod vitamins fork lift carpet diamonds toothpaste opera tickets modem gasoline baby crib cruise headstone wheelbarrow Obviously, recency varies relative to time. Understanding that variance relative to product and product mix is one of the distinguishing skills of successful marketers. Embracing recency as a partial component of the formulaic nature of direct marketing, and indeed general marketing, requires some acceptable scheme of categorization of customers. There must be a measuring device and terminology to describe the state of recency in order for that information to be relevant. There must also be a device for containing and organizing recency information so that specific recency marketing applications can be easily and accurately accomplished. The containment and organizing device is the customer database and the categorization scheme is segmentation. Recency and the Customer Database Nothing concerning customer database technology is mystical or arcane. There are those who would have the database seem mystical and arcane; they have a vested interest in keeping the technology mystical and arcane; in reality it is nothing more than a shoebox. Years ago, peddlers scribbled customer notes on 3 x 5 cards and filed them alphabetically by customer name in shoeboxes. Every purchase was recorded including the date, the product, the amount paid and supporting personal details to trigger future sales knowledge. Every time a customer bought, the card was pulled, updated and refiled in the shoebox. Before a call was made on the customer, the card was referred to and the knowledge brought to the front of the crafty peddler’s memory. Children’s names, color preferences, sizes, birthdays, returns, income and all manner of useful information was recorded to assist the peddler in making another sale in the future. The hearts and souls of the customers were contained in those shoeboxes. When you get down to the practical core, the computerized customer database is essentially file cards in a shoebox. And having gone through the mainframe-mini computer-PC evolution, today’s PC-based customer database can be maintained in about the same amount of physical space as a shoebox or two. But a million customers can be housed in enormous, relational cyber memory. For all of our sophistication and technological advancement we still scribble notations on the customer cards whenever we make a sales call and whenever customers purchase something from us. And, basically, we are interested in much the same information that the nineteenth century peddler captured. The method of organizing the customer database is, essentially, alphabetical by customer name. We use phone numbers, customer numbers, and other identification notations, but we find a single customer through sorting in one way or another. The end result is still the same: We obtain the information we need to make another sale. One component of that information is recency. The peddler, preparing the wagon for an annual trip, would go through the customer cards and redistribute the cards into some logical scheme that ordered the trip and the information about customers. Perhaps only those customers who had purchased every year for the last five years would be called on during that year’s trip; perhaps only those customers who purchased something two years ago but had not purchased anything during last year’s trip would be called on. If the peddler was trying to expand the business, perhaps customers who had bought every year for the last five years would be called on and customers who had bought three years ago but had not bought in either of the last two trips would be visited. Any number of choices were available to the peddler depending on the business strategy and accurate knowledge of customer history as contained in the shoebox database. Once the peddler had selected the customers to be visited, the cards could be put into traveling order. Towns and villages would be scheduled and allocated against the available number of days and the time necessary to make the trip. Here and there, customers would be added or dropped to make the route efficient. No successful peddler ever traveled a random route trusting to serendipity or luck; the marketing plan was studied, calculated and predictable just as it is today. The peddler related customer purchases and trips. Over time, patterns emerged that could be relied upon. One customer might buy every trip no matter when that trip took place. Another customer bought only when the trip was in the fall of the year after the crop was sold and cash was on hand. A third customer bought only every other trip no matter the season and a fourth bought only every other trip in the autumn; some customers were simply totally unpredictable. The intuitive knowledge of these customer patterns was built into the customer cards and contained in the shoebox database. The peddler’s objective was no different from the objective of the database marketer of today: to optimize sales and to generate the greatest possible profit with the least possible expenditure of resources. The shoebox system accomplished that objective then and the computer database system accomplishes that same objective today. Key to the success of the shoebox and the database, however, is the recognition that the primary (indeed, only) purpose was and is marketing. The customer database exists first, foremost and exclusively to create additional sales and to keep customers buying. It does not exist for the convenience of accountants, for generating invoices, for managing inventory or for any of the other bastardized purposes it has been subjected to since its simplistic inception as a sales aid to peddlers. Numerous beneficial by-products of database technology can and do exist, but they exist secondarily to the primary purpose: marketing. Of all the basic tenets of database application, this is the most difficult concept to ensure. Too many diversions are possible once information begins to be organized; too many interests become involved; too many non revenue-producing activities feast as parasites on the database information. CEOs who embrace a formulaic approach to marketing, supported by the organizational and containment attributes of the computerized customer database, must have an unwavering allegiance to marketing to their customers and must eschew all other non selling "noise" emanating from the depths of the database system. A language exists that describes customer recency. The first two branches of that language describe customers in broad terms. Customers are either active or inactive. An active customer is one who is buying; an inactive customer is one who is not buying. Taken further, an active customer is one who has bought at least once in the "recent" past. The definition of recent past is highly variable from business to business. Some businesses define recent past as one year, some two years and some three years or longer. At some point, a decision has to be made that defines an active customer. The continuity of analyses as well as continuity of the marketing plan demands that one works with oranges and oranges. As an example, assume that a company decides an active customer is one that has purchased at least once in the past three years. Any customer who has not purchased something in the past 36 months is, then, an inactive customer. It may stretch the imagination to classify a customer as active with only one purchase in 36 months, but there are various reasons that this is done. The value of the active customer list is often the major asset of a direct marketing company; by claiming active customers three years back, the total active customer count increases and the value of the list goes up. Experienced marketers recognize this ploy, but inexperienced people acquiring direct marketing companies may not understand this aspect of asset value. Similarly, customer list rentals depend on active customer names. If the length of time defining an active customer is shortened, the number of older names that can be rented decreases and the company’s Other Income line is lessened; if the length of time defining an active customer is lengthened, however, the total active customer count increases and the rental income likewise increases, albeit only for a short period of time until the renters figure out that the names are old. This is again a ploy that an experienced marketer will quickly see. Also, the element of denial exists in the classification of active and inactive customers. If a company is slowly expiring and the CEO cannot bear to accept the physical evidence of the demise, the definition of active customer will stretch back in time until the CEO reaches the "denial comfort level." This is, of course, ultimately fatal. For a well-rounded appreciation of recency, the image of the customer base as a reservoir is often helpful. As long as new water pours into that reservoir, it stays full. But if water is pouring out at the discharge end and no new water is coming in at the intake end, the reservoir will go dry in a hurry. That is exactly what is happening with recency in the customer base. If customers do not buy a second or third time or more, they are not refilling the reservoir. If, additionally, new first-time customers are not pouring in at the intake end of the business, then the business is going dry. Customers "fall off the back end" constantly; that is normal, unavoidable customer attrition. They move, die, get mad and go away, go bankrupt, close up shop and any number of other things that cause them to fall off the back end. If one customer a day falls off, then one customer a day must be added just to maintain the customer base. If two customers fall off the back end and only one comes in, then the business will soon lose 50% of its customers, a condition generally known as death. Almost all marketing effort is done to ward off death. Obviously, it is essential to know the exact attrition rate that is taking place at any moment in the business if you are to avoid the Grim Reaper. In part, that is discovered through recency segmentation. Customers are classified as one-year, two-year, three-year, four-year or five-year buyers. These categories are totally arbitrary; they could just as easily be quarterly, 6-month, 12-month, 18-month, or 30-day, 60-day, 90-day, or any other division and category of time that makes logical sense for the particular business. For continuity, one- through five-year segmentation categories will be used for all examples. The one-year buyer segment of the customer database is, in reality, a small box inside the larger shoebox. Inside this smaller box are all the 3 x 5 cards of all the customers who have bought at any time in the last 12 months. If there has been a purchase of any kind in the past 365 days, those cards are updated and filed away in the little one-year box. That way, when the marketer wants to refresh the memory as to who has bought something in the last year, the little box can be lifted out of the larger shoebox, opened up and all the customer index cards can be reviewed. The two-year buyer segment of the customer database is, similarly, another small box inside the shoebox. It contains all the cards of customers who bought up to two years ago but who have not bought anything in the last 12 months (because all of those people have been moved into the one-year buyer box). Another way to state this is the two-year segment contains all customers who have bought in the past 24 months but not in the past 12 months. If one of those customers who has not been heard from in the past 12 months calls and buys something today, the little card will be updated and moved into the one-year buyer box. No record of this customer will remain in the two-year buyer box because the customer has reaffirmed loyalty and has been promoted to the most recent segment, the one-year box. The three-year buyer segment is a third box inside the shoebox and it contains all of the cards of customers who bought up to three years ago but who have not bought in the past 24 months. In other words, these are customers that have not been heard from in two years, although they did buy 25 to 36 months ago. If one of these customers called and bought today, the card would be updated and placed in the little one-year box. Again, no record would be kept in the three-year box. The four-year buyer segment is a fourth box in the shoebox. Contained inside, all in order, are the customers who bought something four years ago but have not bought anything in the most recent 36 months. Their last purchase was made sometime between 37 and 48 months ago; they have made no purchase in the last 36 months. And the five-year buyer segment is the fifth and last box inside the shoebox. The customers in this collection bought something 49 to 60 months ago in the dark and dim past but have not bought again in the last 48 months. Generally, a sixth box exists. It contains all the customers that have not been heard from since 61 months back and beyond into the dim and receding mists of time. This box often has customers that bought one time ten years ago and never bought again. The dust is usually extremely thick on this box partly because there is so little activity and partly because so few marketers open up the box and dump nonbuying customers from the other boxes into it. From time to time, however, this little treasure chest is trotted out on parade disguised as a corporate asset. Go figure. Now the minute a customer from any of the five or six boxes makes a purchase, that customer is automatically reclassified as a one-year buyer. They have bought in the most recent 12 months and are, therefore, the most recent customers. Their cards are rooted out of their previous box and added to the one-year file box. From just this rudimentary knowledge about recency, one or two insightful conclusions can be drawn. If all of the customers are in the one-year box, the company is a real barn burner. But, if all the customers are in the five-year box, the company has apparently been out of business for the last four years and nobody has told them. In reality, we can conclude that most companies have some logical and normal distribution pattern of customers in all of the recency boxes. The difference between great companies and not-so-great companies is in the shift of the weight of the pattern. Mediocre companies have too few customers in the one-year box and too many in the two, three, four and five-year boxes. In other words, more customers are falling off the back end than are coming in on the front end. The reservoir is drying up. It is important to recognize that each of the boxes inside the bigger box is constantly moving forward in time. If today is April 1, then the one-year box will extend back to March 31 one year ago; tomorrow, the box will be between April 2 and April 1 a year ago. Each of the other yearly boxes is moving in time as well, as is the whole shoebox. In other words, the customer segmentation database is dynamic; it never stands still in time. It is also important to recognize that individual buying customers can only move one direction: forward in recency. A one-year customer can never move to the two-year box as long as a purchase is made sometime in the last 365 days. A one-year customer can move to the two-year box provided that no purchase has been made in the past 365 days. In other words, the act of buying is the catalyst for recency. Combining these two recognitions, then, leads to the logical conclusion that a customer must make a repeat purchase in some acceptable length of time for the company to be profitable. That logic — simple as it may be — is the heart and soul of direct marketing. Recency velocity is the speed at which one-year recency is achieved. As an example, a daily buyer of The Wall Street Journal has a recency velocity speed of 1, the fastest velocity possible. A buyer of house paint may have a recency velocity speed of 1,460, a slow velocity representing 1,460 days since the last purchase, or four years. If a customer purchases five times a year, the recency velocity speed is 73, a rate of one-year recency reachievement every 73 days. Customers have different recency velocity speeds. One customer may be a 110, another a 56, and a third a 250. In the aggregate, all customers combine to determine an average recency velocity speed. The three customers above have an average recency velocity speed of 138.7. Another way of saying this is that the average customer reorders 2.63 times a year, about every 139 days. Recency velocity is an important concept because it describes the dynamism of the individual and the average customer. In an ideal world, the recency velocity speed will constantly increase; that is, it will speed up for both the individual customer and all customers as a whole. The more the velocity increases, the more orders there are during the year. If you can move individual customers closer to a velocity of 1, then the whole customer base will move closer to daily ordering and all of the cards will be in the one-year recency box. That is a good thing. If, however, the recency velocity speed for individual customers and for all customers on average is slowing down, then the company is experiencing decreasing dynamism; the business is slowing down or the reservoir is drying up. That is not a good thing. Recency momentum is the force of the change in recency velocity. If recency velocity is improving from 130 days to 120 days, the recency momentum is a +10. Conversely, if recency velocity is deteriorating from 120 days to 130 days, the recency momentum is a -10. Obviously, a positive momentum is a good thing and a negative momentum is a bad thing. If the peddler goes out selling along the route, the recency velocity is going to speed up as customer after customer buys. The recency velocity average for all customers will also speed up even though many customers are not being called on during that trip. For a brief period — the length of the trip — the recency momentum will increase positively. When the trip is over, the recency momentum will decrease negatively. The same experience occurs when direct marketers mail out an offer. Both the recency velocity and the overall recency momentum increase as orders flow in; both decrease as the mailing ends. It is the ebb and flow of recency velocity and recency momentum that is significant. If the recency velocity speed is faster with each successive offer, then the overall recency momentum will achieve successively higher and positive levels. The business is in the pink of recency health. Over long periods of time, say 36 rolling quarters, the ebb and flow of recency velocity and recency momentum should slope upward when displayed on a chart. Only a sufficiently long view will allow for the smoothing of individual marketing efforts and give a true picture of the state of customer recency. While the business health is determined by the macro-velocity and momentum of recency, a more important micro-velocity and momentum state of health must be diagnosed for the individual customer. The ability to see the speed and force of recency velocity and recency momentum for the individual customer is a key factor in database marketing and in the analysis of recency, frequency and monetary value. It has been said that understanding a direct marketing company begins and ends with understanding a single customer thousands of times over. If you have the luxury of individual customer knowledge, you have the luxury of understanding your business to the same degree as the peddler of old. Unfortunately, few direct marketers have that knowledge and, as a consequence, operate their businesses with an obsolete and irrelevant "mass" marketing strategy. But what if it was possible to look at recency velocity and recency momentum for each individual customer? What if it was possible to create a recency velocity history and predict when that customer was most likely to buy again? And what if unique efforts were directed to that one customer and both recency velocity and recency momentum were improved? Concrete, historical knowledge at the individual customer level, repeated thousands of times over, produces intimacy. Intimate knowledge, built up through information about the buying behavior of every customer, is the foremost safeguard for direct marketing success. A huge gulf exists between having intimate knowledge and being able to have intimate knowledge. With the sophistication of database technology, almost every direct marketing business has the ability to accumulate information; few companies actually do something with that information. All too often, mountains of printouts exist that contain the answers to every conceivable question, but no one looks at them. Individual customer intimacy entails study. Customers are tracked, reviewed and interviewed; their purchases are questioned and motivation is sought; certain customers are returned to time and time again to determine shifts in attitude, needs and opinions. CEOs take note: If you do not know the individual customers’ reasons for positive or negative changes in recency momentum, you are flying by the seat of your pants and headed for a huge mountain top that you can’t see in the fog. At any given moment in time, a precise number of customers exists. This number is the sum of all of the customers in all of the various recency segmentation boxes inside the big shoebox. Each day, the number of customers changes and customers move between the recency segmentation boxes. As an example, a customer belonging to the three-year recency segment–who has not been heard from in the past 24 months up through the past 36 months–makes a purchase today. That customer moves from the three-year box to the one-year box, a loss of one customer from the three-year segment and a gain of one customer in the one-year segment. At the end of today, the one-year recency segmentation box will have added all of the customers who bought today from the two-year, three-year, four-year, five-year-and-older boxes. At the same time, all of the first-time buyers or new customers that have bought today will have been added to the one-year file. All of the customers who aged beyond one-year without buying as of today will have been moved to the two-year box. The one-year box has a net gain and a net loss, and every customer can be accounted for through a reconciliation. At the end of today, all of the customers in the two-year file who purchased today will have been moved to the one-year box. And all of the customers who have aged beyond 24 months without buying will have been moved to the three-year box. The two-year box also has a net gain and a net loss; customers moved either to the one-year box or the three-year box; no customer "leakage" or "shrink" can be allowed with a proper reconciliation. The three-year box, at the end of today, will lose customers who purchased to the one-year box. Customers who have aged to 48 months, as of today, without buying will have been moved to the four-year box. Again, a reconciliation accounts for every customer. The four-year box sends customers to the one-year box if they bought today and to the five-year box if they aged to 60 months without buying. The net gain and net loss can be accounted for down to the individual customer. The five-year-and-older box also may have sent a few customers to the one-year box, but mostly it receives a steady stream of aging customers from the four-year box. This is a terminal box. It doesn’t send any customers further down the aging line because this is the end of the line. All of the old customers who bought a long time ago and stopped buying are in residence in the five-year-and- older box (or the six-year-and-older box or however many boxes your company maintains). By calculating the net gain and loss in each of the segmentation boxes, an accurate reconciliation of the customer count is accomplished. If the day began with 100,000 customers in total and 200 new customers were added for the first time, the customer count at the end of the day will be 100,200. The movement between the boxes does not alter that straightforward count; getting all of the customers properly in their boxes is, however, a fairly complicated procedure. Beyond the total number of customers and the need for accuracy in keeping them properly counted, there are a number of interesting things revealed by the reconciliation process. If, suddenly, there is no movement of customers from the two-, three-, four- and five-year boxes into the one-year box, something is seriously wrong. The customer base has stopped repeat buying. On the other hand, if all of the movement is from the older recency boxes into the one-year box and no new customers are being added to the one-year box, new customers have stopped entering the business, another serious problem. Over time, the reconciliation process can track a large concentration of customers. Suppose that you mailed out an introductory offer for a product at 50% off. As a result, 5,000 new customers purchased once but never bought again. This group of 5,000 will appear in every reconciliation as they move out of the one-year box, into the two-year box, then into the three-year, the four-year and ultimately the five-year box. Being able to spot large boluses of customers in the recency data is important when evaluating a company for acquisition or when reviewing the quality of the marketing history and the quality of the customer base. Without recency segmentation or history, these 5,000 questionable customers would simply be a hidden part of the overall customer count. The intricacies of recency reconciliation are many. When marketing efforts are directed specifically to the two-year, three-year, four-year and five-year customers, the response rates for each group should be relatively consistent. If the customer count in the three-year box suddenly drops, only three possibilities exist to explain the activity. The first is that a large number of three-year customers have suddenly purchased. This can be verified through reconciliation by confirming that the one-year box increased in count by a like amount. The second is that a large number of three-year customers aged as a group and dropped into the four-year box. This can be verified through reconciliation by confirming that the four-year box increased in count by a like number. The third possibility is that an error in recency reconciliation has occurred, a common problem when tracking large numbers of customers. The percentage of customers moving to the one-year box compared to the percentage moving to the next box in the aging process is another interesting intricacy. If the three-year customers, over time, average a 6% movement to the one-year box and a 30% movement to the four-year box, what does it mean when one month 40% of the customers show up in the one-year box? Obviously, there was a significant marketing effort designed to reactivate three-year customers. A change in the average rate of movement, either forward or backward in recency, is an indication that something is going on in the business that you should know about. The only way to get this information is to track the changes in movement in each recency category every month over time and to look for trends in customer activity. To assure accuracy in analyzing trends, the accuracy of the reconciliation is essential. The recency boxes and their individual and combined reconciliations are no different than a checkbook. At the end of every month, the checkbook must balance to the penny; at the end of every month the recency figures must balance to the customer. Any discrepancy in the reconciliation must be explained and corrected. Customers do not either appear out of thin air or disappear into the nothingness of vapor. They went somewhere and that movement must be accounted for precisely. Under the arguable assumption that a one-year recency customer is a better customer than a two-year or three-year recency customer, recency segmentation is an essential pillar of list rentals (your customer list being rented to other companies) and list renting (other companies’ customer lists being rented by you). When all is said and done in these tortuous economic times, list rentals can often contribute to the bulk of a company’s profits. One of the attractions of direct marketing is the income that results from list rentals, income that is, essentially, "found money." A company having 500,000 active customers can rent those names for, say, $80 a thousand. If all 500,000 are rented, the company earns $90 x 500, or $40,000. If the entire list is rented 50 times a year, the gross rental revenues are $2,000,000. Renters of customer lists are interested in the quality of potential response. Therefore, a list of one-year recency customers is more desirable than a list of four-year recency customers. If the one-year list rents for $90 per thousand, the four-year list may bring only $60 a thousand or less. Typically, renters pay top rental dollar only for the most recent quarter of recency, the "hot" names, those customers who have purchased in the past 90 days. The "first quarter" recency names can bring as much as $110 per thousand in rental income. The next group — the second quarter names — may rent for $90 and the third and fourth quarter names may bring $80 per thousand. For list rental purposes, quarterly recency segmentation is essential. Think of this as a further segmentation of the one-year shoebox, dividing it into quarterly subsections with the customers arranged and moved accordingly. Ideally, the perfect company is one that has all of its customers in the one-year box and — inside that box — all of the customers in the current quarter segment of the year. Of course, that does not happen, but it’s nice to contemplate such a perfect money machine. As a list owner, you would like to rent all of your names — often; as a list renter, you want only the names of another company’s most recent buyers. The dichotomy is clear. The common requirement, however, is that the names must be segmented by recency at the minimum. When renting names by recency, the most common request is for the "quarterly hots," those customers who have bought in the last 90 days. Experienced direct marketers will also rent quarterly hot lists from one, two or three quarters back. The logic surrounding this recency strategy in rentals is that every company under the sun is renting the current quarter names and those customers will be bombarded with mailings. Why not mail to equally good names that are two, three or four quarters back? The assumption is that they will not receive as much mail and your offer will stand a better chance of surviving the clutter. There is wisdom in this strategy, and for some products or services it works well; for others it doesn’t. Again, the necessity for recency segmentation is the one constant, whether you are renting names to others or are renting names from others. Recency is but one element of an overall list rental strategy; many other segmentations and attributes of customer lists are taken into consideration when formulating a list plan. In most instances, the initial cut is made on recency. A recently active customer has the greatest potential and that potential diminishes over time. Rarely do other marketers specify rental of the six-year- and-older treasure trove of customers who have not bought anything in the last five years. The percentage exists in the current customers. Recency and the Number of Customer Mailings or Contacts Thinking again as a peddler, the question of how many trips a year to make to the customers is a basic and important question. It costs money to "make a trip" whether that trip is by wagon, mail or telephone. If every customer bought on the annual trip, the peddler would ask sooner or later, "What if I made two trips a year?" If every customer bought on each of the two trips, the question would become, "What if I made three trips a year?" If the customers all bought on each of the three trips, the question would be, "What if I made four trips a year?" As long as enough customers buy on each trip so that a profit is produced, the number of trips can increase. Direct marketers ask the same question, only phrased as "How many catalogs (or solos) can I mail to a customer every year?" The answer is the same: As long as enough customers buy on each mailing so that a profit is produced, the number of mailings can increase. Whether you are a peddler or a cataloger, the justification process is the same. You will market to those who buy from you regularly. And, if you are smart, you will market to those customers more often, up to the point where it no longer is profitable to do so. In the early days of direct marketing, catalogers would send out one catalog a year. That grew with positive experience until four catalogs a year were being sent to active customers. Today, the number of catalogs being mailed by business to business as well as consumer marketers can be as high as 26 a year, one every two weeks. Unless those marketers are either really stupid or really self-destructive, it’s a pretty safe assumption that they are making money on that number of mailings. Not every customer receives the maximum number of mailings or contacts a year, only those who respond with regular purchases, only the top of the one-year customer segment. The best customers get the most attention; the others get less. If the one-year customer receives 26 catalogs, the two-year customer may receive only 18 mailings. Three-year customers may receive only 12 catalogs; four-year customers, 8; five-year customers, 4 issues. A hierarchy of quantity of mailings exists based on the profitability of the customer response. The rule of thumb is "You can mail a customer one more mailing as long as that mailing is more profitable than the last prospect mailing." At the point where another customer mailing is less profitable than the worst prospect mailing, you have reached maximum customer mailing quantity. To illustrate this, suppose that a choice exists between mailing the one-year customer segment catalog number 18 for the year or mailing the dollar equivalent in prospect catalogs. As long as the added customer mailing makes more money than the prospect mailing, you can do it; when it fails to earn as much as a prospect mailing, you have reached saturation. Testing the two-year, three-year, four-year and five-year segments in the same way may allow the addition of profitable customer mailings and ratchet up the whole customer marketing effort. Once again, strategy is founded upon the element of recency. Aside from pure profitability as the deciding factor determining the number of mailings to customers, the new customer acquisition strategy must be considered. It does little good to overmail to customers at the expense of bringing new customers into the company. Without new customers the reservoir will run dry even with great buying from existing customers. Unless an aggressive approach to new customer prospecting is maintained and carefully evaluated, all of the mailing to customers will not stave off eventual decline. Generally, if a choice has to be made between active prospecting for new and fresh customers or mailing to existing customers in the older recency boxes, the strategy of active prospect mailing makes better long-term sense. So much depends on the definition a direct marketing company uses to describe a customer. If a customer is someone who has not bought in four years and if that customer is receiving 18 catalogs a year, there is probably something seriously wrong with the perception of customer qualification by management. Direct marketers are often unsure how to treat inquiries. Should people who request information be treated as customers, or should they be treated as one-time prospects until they actually make a purchase? Different strategies abound in the industry but several common sense points can be made. An inquiry is certainly more recent and more qualified than a complete unknown. The act of making an inquiry has qualified that person to some extent, at least more so than a name on a prospect list. An inquiry should probably be offered the opportunity to purchase for some period of time before discarding that name or relegating it to the "dead" file. How long mailings should be continued without a purchase being made is the question. In some companies, inquiry names are simply added to the one-year recency file and mailed as one-year customers. Often, six months of regular and frequent mailings are made and the inquiry is then removed from the one-year file and either discarded or sent down to the box marked "Others." More aggressive companies give the inquiry only one or two mailings to make a purchase before sending it down. It all depends upon the performance of inquiries for a particular company. Some products take forever to win a buying decision; some take 30 seconds. One company may have to mail an inquiry 50 times before the customer actually buys; another may mail only once. The key is to track and measure the response of inquiries and compare that response with other categories of recency and other categories of prospect mailing. The analogy is much like the tenacious sales call: Keep calling and keep asking for the order until they buy; of course, there is a point where common sense and profitability rule. Generally, in business to business direct marketing the length of time an inquiry is given to make a purchase may be less than in consumer direct marketing. If, after a reasonable period of time and mailings, the business customer has not purchased, the inquiry can be returned to the great pool of prospects. One strategy is to maintain a list of inquiries and add them to routine prospect mailings under the assumption that the nonbuying inquiry is still a better bet than a total unknown. If inquiry names are maintained as a separate box in the customer database, then moving buyers to the one-year recency box has to be factored into the database schematic. The inquiry box simply becomes another stop on the reconciliation route. For those new to the direct marketing business, it can often be revealing to find out how inquiries are handled. If inquiries are separately categorized and aggressively deleted, well and good; if inquiries are added to the one-year customer recency segment and then aged through the other recency years as if they are buying customers, something is amiss. One way to bolster the customer list (and the company assets) is to add inquiries as actual customers. This makes the business look good and contributes to list rental income. One would not want to purchase a company consisting of mostly inquiries however. Recency and Customer Attrition
Rates If you get 100 new customers today and 70 of them reorder within the first year, the one-year recency reorder rate is 70% and the attrition rate is 30%. That is very straightforward; no special incantations are required to understand that fact. But what is the attrition rate of customers in direct marketing? That is a question that is asked far more often than it is answered. In all of business there exist thousands of maxims, those little rules of thumb that everybody knows. One of those maxims is that you turn the entire customer base once every six years. In other words, if you have 100 customers today, six years from now you will have an entirely different 100 customers. The original customer will be gone and will have been replaced by new customers. Whether this is true is debatable. But some part of this maxim is true. You gain and lose customers constantly and at some relatively stable rate barring disasters. As an example of recency and attrition, a purely hypothetical example may be instructive. For some companies this model will be representative; for others it won’t be, but the principles are what we are after. Today the company receives 100 first-time orders. These are new customers and are placed in the one-year recency box. Over the next year, 50% or 50 of those customers reorder and are retained in the one-year recency box. The other 50 customers are never heard from. The one-year attrition rate is 50%. During the second year, 13 (or 25% rounded) of the 50 original customers still in the one-year recency box never order again; 37 do order again. The second year attrition rate is 25% of the original 50 new customers who ordered again. During the third year, another 25% of the remaining 37 customers disappear, leaving 28 of the original customers in the one-year recency box. During the fourth year, 25% of the 28 customers, or 7, evaporate, leaving 21 original customers in the one-year recency box. During the fifth year, 25% of those 21 customers, or 5 more, go away forever, leaving 16 of the original customers to carry on. In the sixth year, another 25% will dissolve, leaving 12 of the original 100 customers still active in the one-year recency box. A total of 88 customers will have aged through the two-year, three-year, four-year and five-year recency boxes to arrive in a lump in the six-year-and-older box. In this example, the attrition rate is 50% the first year and 25% of the remainder in each succeeding year. The bottom line is that only 12% of the customers continue after six years. Another way of looking at this is that you have to replace 88% of the customers every six years just to stay even. Attrition rates vary from company to company, industry to industry, and from marketing strategy to marketing strategy. What does not vary, however, is the fact that attrition is a dynamic process; it cannot be denied and it does not go away when ignored. The glaring recognition to be etched on the marketer’s mind is that unless new customers arrive at an incoming rate greater than the outgoing attrition rate, the company is going out of business every day. To simply stay even, you have to add a minimum of 15% new customers on average in each of the six years. Clearly, the ability to know the exact attrition rate in each year of recency on a monthly, quarterly, annual, and 20-rolling-quarters basis is essential for any direct marketing company. Without this knowledge the company is flying blind and by the seat of its pants. More important is the knowledge of the change in the attrition rate and the trend of the change. Only three possibilities exist for attrition: improving, getting worse, or remaining the same. There are no other choices. A measurement of the effectiveness of marketing to individual segments of recency, such as two-year versus three-year, is the change in the rate of attrition in one segment over another. If the four-year recency segment marketing program is performing better than the two-year recency segment, relative to attrition, then something can be learned from that fact. Remember, the ultimate goal is to move all customers into the one-year box. Recency and Attrition
Recovery During the first year 50% of the customers never order again. That is an enormous waste. What if 50% of those customers could be recovered and kept in the first-year recency box to buy again another day? That question is the heart of the concept of customer retention. If, instead of losing 50% of the new customers in the first year, the company lost only 25%, the end result would be a 50% gain in customer retention at the sixth year; 18 customers would remain instead of the 12 in the example above. That is a good thing. But there is a cost associated with retaining those additional customers. If the cost to keep six additional customers is greater than the profitability gained from those six customers, then customer retention may not be such a good thing. The only way to know the answer is to know the costs. Alas, in the new order of direct marketing even accountants have a place. Our earlier book, Libey On Customers, explores customer retention strategies and techniques in great detail. The bottom line is that customer retention is the direct marketing survival strategy for the future and you simply cannot afford to lose a single customer. Companies that are not obsessively focused on retaining customers are fast approaching extinction. Everything of importance in this business begins and ends with the individual customer; retaining them is the prime directive. Recency and Multiple Customers Keeping customers in their correct boxes is complicated. For business to business and some consumer direct marketers a decision has to be made about who is the customer. That simple decision can complicate the recency segmentation beyond belief. Who was the customer of the peddler? Was it the mother, the father, the children, or was it the family unit, the farm itself? Did the peddler decide to include a farm on the route based on which family member purchased last? How did the peddler manipulate the 3 x 5 cards when several members of the same farm family were the customers but bought on different trips? A business is an entity. It continues usually without regard for the people who are the employees. Employees come and employees go; the business is the stable entity. If a business has several employees who are customers, which is the customer of record, the individual people or the business entity? The answer to that question has a lot to do with how the recency segmentation boxes are set up and handled. The peddler would have noted the purchases from each family member on the same card. The family unit itself would have been the customer of record. It made little difference to the peddler who bought as long as someone bought on each trip. The farm was the target. But does that work in direct marketing? If an employee is the customer, does the business and the individual employee continue to be carried in the one-year box or only the employee or only the business? If 50 employees from the same company are customers, do all 50 receive mailings even though only 2 have bought anything in the last 3 years? It gets complicated, doesn’t it? Look at the numbers. Suppose that you are mailing at the rate of 26 catalogs a year to one-year recency customers. If it is your policy to mail each customer — a contact — inside the company and you have 20 contacts for one company, you will mail 520 catalogs into that entity. If, in addition, you also mail the entity, you will add another 26 mailings for a total of 546 catalogs a year. But what happens when only 3 of those 20 contacts in that single entity are in the one-year recency box? Maybe there is only justification for mailing 26 catalogs to those 3 contacts. This would bring the mailings down to 78 catalogs plus 26 for the company itself, or a total of 104. The difference in policy is a reduction of 442 catalog mailings into that company. At $.80 each, this mounts up to $353.60 a year, a rather significant cost. The answer, of course, is that customers are people first and last, and you treat the individual person as the customer. The individual person has to meet the recency standards to continue to be treated as a one-year recency customer. The individual contacts inside a company are mailed based on recency performance. Whether the entity is mailed as well is a policy decision. There are those who believe that sending the company a mailing assures that someone will receive it; there are those who mail only contacts and not the entity. The decision is determined by performance. If the entity mailings produce acceptable sales, continue mailing the business. If only the contact mailings produce sales, discontinue the entity mailings. Again, it is not a simple matter of yes or no but a matter of doing the tracking and the evaluation of the performance of the mailings. The complicating factor is in the structure of the database necessary for tracking multiple customer contacts within a single company entity while keeping the individual contact recency segmentation accurate. At the same time, you do not want to lose the "velocity" of recency that the company itself is experiencing through its employee customers. There is no right answer. The structure is one based on policy and on the capacity of the database program and its infinite flexibility. Recency and Why It Isn’t Enough For all of the supposed sophistication of North American direct marketing companies there is an appalling lack of sophistication in the use of recency, frequency and monetary value analyses as an integrated analytic tool. Many companies are competent with recency alone, but are as yet unable to integrate frequency and monetary value. For those companies existing with only recency as their formulaic touchstone, they are staring into the eye of disaster. The reason for this lack of sophistication is rooted in the attitudes, the capital resources and the database technologies of many direct marketing companies. Believe it or not, these companies are still managed with the same marketing attitude that was prevalent in the 1970s and 1980s: Put a lot of stuff in the mail and people send you money. Often these companies don’t have the money to move from an inventory- tracking, pick-ticket-producing, invoice- printing database to a true marketing database with just the rudiments of recency, frequency and monetary value analyses, let alone the advanced analytic capabilities. The cost of hardware and software to put an efficient, functional and relatively up-to-date relational database in place is prohibitive. As a result, they continue to spend chunks of money on retrofitting old databases in the hopes that magic will occur. Nothing is sadder than a CEO who has made an enormous error in choosing database technology having to return year after year to the board of directors for ever-increasing capital financing to shore up an ill-conceived, obsolete and non-functional computer system. The tap-dancing, promises, and constant excuses are pathetic if not irresponsible. At the same time, the company is slowly disappearing because nobody has a clue about what is happening and the poor CEO is hoping to make it through another year to retirement, take a cash out or find a whole new career. Recency alone does not a direct marketing titan make. Recency only gives one-third of the essential, base information for managing a direct marketing strategy. On top of recency must come frequency, monetary value and a host of other information about the individual and collective customers. Direct marketing companies operating with only recency segmentation are analogous to a golfer playing a PGA course with only a seven iron. It’s great for about one shot out of twenty, but you can’t win with it in competition with other golfers having a full bag of clubs. Knowing when a customer bought is great for a mailing strategy, but it does nothing to answer the important questions like, "How often does a customer buy?" or "How much money does a customer spend?" or "Where do the best customers come from?" or "What do the best customers buy?" Basically, recency allows you to choose which customers you will visit and how often you will visit them. Recency also tells you something about the individual customer purchasing trend and, collectively, the purchasing trends within the entire customer base over time. And, through the reconciliation process, recency is an indicator of growth and shrinkage, of health and disease, the pulse of the formulaic direct marketing company.
Real-Life Frequency Characteristics Frequency is how many times a customer buys. It is an indicator of usage and satisfaction. As an aggregate of all customers, it is a primary measurement of demand. By itself, it is the second important indicator of the cardinal three — recency, frequency and monetary value (RFM). Frequency describes a state of a customer’s demand level. Broadly described, demand for products and services can be described as high, average and low. A customer who purchases 12 times a year could be considered to be a high demand, high frequency customer. Another customer who buys four times a year may be an average- frequency, average-demand customer. And a customer who buys only once a year may be thought of as a low-frequency, low- demand customer. All three would, however, be considered active, good customers as they consistently affirm their loyalty by regenerating their one-year recency classification. From purely a frequency point of view, a customer who buys 12 times a year is more valuable than a customer who buys only once a year. Twelve opportunities to engage the customer exist rather than just one. Every interaction is, potentially, an opportunity to strengthen the loyalty, the bond, the reliance and the retention percentage between the customer and the company. Also, 12 times the number of opportunities to sell other products or services are available with the high-frequency customer. If you think of frequency as the number of times a customer walks into your store, you will understand the importance of frequent customer contact. But, there is a great difference between walking in and buying something. For the direct marketer, frequency entails the act of buying. The more frequent the purchases, the higher up on the frequency hierarchy the customer rises. For continuity program marketers (Book-of-the-Month Club), a 12-purchase frequency or more during the year is desired and those customers who attain that level of frequency are prized. Frequency is influenced by product and usage. The frequency of an automobile purchase is much less and entirely different from the frequency of orders for checks. An automobile dealer may have an average customer with a four-year recency and a one-time frequency; a check printer likely has an average customer with a one-year recency and a two- or three-time frequency. At the extreme, a bridge building company may have customers with a 100 year recency and a one-time frequency. Generally, nondurable or consumable products produce higher frequency than durable products. The comparison of bridges and checks makes this concept instantly clear. The faster the usage or deterioration rate, the higher the frequency. Products having a shorter shelf life tend to drive a higher frequency of purchase; milk is an example. Consumable products can be highly seasonal and frequency can be correspondingly seasonal. Consider ice- melting products, for example. Essentially, sidewalks, water, and freezing temperatures are necessary to drive purchases of ice-melting products. That implies a seasonality factor and it is logical to assume that multiple purchases (a higher frequency) occur during the winter season. The customer may be a one-year recency customer, but the frequency may be compressed into the winter months. Frequency can also be influenced by deviations in external forces. An abnormally severe and wet winter can increase the frequency of ice-melting products purchases beyond that experienced during an average winter. Similarly, the frequency of thousands of products was skewed dramatically in the midwestern states following the devastating flooding in the summer of 1993. By and large, commodity products have a steady, fairly predictable rate of frequency. A customer buying business forms for a manufacturing company tends to purchase 2.6 times a year, or about every 140 days. For some reason, business forms inventories are purchased for about four or five months at a time. The average, loyal customer will remain in the one-year recency box and will have a frequency of 2.6 purchases. If the frequency of that customer’s purchases changes, it is an indication that something about that customer’s business is changing, something that the forms marketer must be aware of and must evaluate for action. Frequency is also influenced mightily by fashion and fad. At the height of the athletic shoe fashion craze, some customers may have had a frequency of 12 purchases a year. A new pair of pumpable sneakers every few weeks was a requirement for being in step with the latest fashion. As that fad diminished, the rate of customer frequency diminished as well — rapidly. Virtually overnight, millions of customers went from the one-year recency box and a 12-time frequency to the two-year recency box and a zero frequency. The fad had expired. For a one-product company that is a big problem. Frequency can wax or wane with the vagaries of regulatory or legislative control or deregulation. A product that is suddenly restricted will experience a drop in frequency; a product that is suddenly outlawed will cease frequency altogether. A restricted or banned product that is suddenly deregulated will experience an initial surge in both recency and frequency, moving toward equilibrium afterward. And, of course, frequency is influenced, perhaps to the greatest degree, by satisfaction. If the customer is not satisfied, regardless of the reason, frequency diminishes or stops. Satisfaction comes in many shapes and guises; it is often impossible to know what constitutes satisfaction in the mind of the customer and often it can be irrational. But, for the most part, satisfaction involves basic expectations: price, timeliness, courtesy, attention and quality. Marketers cannot control the irrational, but they can control the basic customer expectations. There is a restaurant in a neighboring town that we try every two years. The service is consistently poor, the food is consistently mediocre and the price is consistently too high. We are a two-year recency customer with a frequency of one time. We will never be regular customers until the satisfaction reaches a fair level of expectation. In our case, it is always a two-year experiment — "Let’s see if it is still as bad as ever." But if satisfaction were delivered, we would be consistent one-year recency customers and maybe a 10 or 12 frequency. The basics are not being met. As you can see, the real-life characteristics of frequency are many and varied. Frequency is not as cut and dried as recency. The complexities of frequency make it much harder to understand the underlying motivations. The numerical count of frequency is the easiest part and is programmable in any basic database; the customer’s hidden reasoning behind purchasing frequency cannot be programmed and, unless carefully researched and analyzed continuously, remains forever a mystery. Theoretically, if product usage increases, frequency increases. This may or may not be true. To explore the complexities of frequency and product usage, we will look at business forms as an example. Business forms are the ideal product to learn more about frequency and product usage because they are demand driven. If a business makes 1 sale, 1 invoice is used; if the same business makes 100 sales, 100 invoices are used. The product usage varies in direct response to demand. Product usage of business forms can be influenced by many things, internal and external. Internally, a company may shift technology from impact printers to laser printers. As a result, plain paper is used to print invoices and preprinted standard or custom invoice purchases will decline to zero. As a result, product usage and frequency drop to zero. Technology has obviated the need for the product. Externally, an economic recession may be responsible for a decline in sales of 50%. As a result, invoicing declines by 50% and the usage rate for invoices and purchasing frequency drop correspondingly. The economy can drive usage in either direction. Internally, cash flow can influence product usage rates and frequency of purchasing. If cash is plentiful, a company may purchase a year’s supply of business forms to take advantage of the price break on large quantity purchases. Conversely, if cash flow is tight the same company may shift to monthly purchases of business forms to conserve cash even though the price of the smaller quantity is higher. Converting cash to a year’s inventory of business forms to sit on a shelf is not in the best interests of a cash-strapped company. Internal purchasing authority can influence product usage and frequency. The person authorizing the purchase may be limited in authority to purchases of $100 or less. Rather than go through the internal justification and requisition process or the competitive bidding process, a quantity of business forms costing under $100 is ordered several times a year. External, uncontrollable events can influence usage and frequency. The Mississippi River basin flooding of 1993 may well result in massive replacements of business forms and dramatic increases in frequency and usage. Every time a zip code or a telephone area code is changed, replacement business form usage increases and purchasing frequency increases from the norms. Clearly, the reality of product usage and purchasing frequency is some blending of both external and internal circumstances. Overall, the economy, cash flow and product maturity cycles tend to be the primary regulators of usage rates and frequency of purchasing. Other important factors enter the picture also, but to a lesser degree. Discount offers can influence frequency; the whim of personal preference has some effect; mergers and acquisitions have influence on frequency; corporate downsizing takes its toll; quality control programs minimize errors and can lower product usage rates and frequency; advertising influences sales; international expansion changes the levels and rates; any number and combination of the entire panoply of commercial influences result in product usage and frequency variances. And these are only a partial listing of the considerations for one product: business forms. Imagine the complexities driving usage and frequency for the millions of products that exist globally. Little wonder that so few companies have any understanding whatsoever of what lies at the foundation of frequency.
Price has an effect on frequency outside of the dominant influences of the economy and cash flow; price defines competition. All things being equal, the customer will buy on price. This was not always true, but increasingly price has become the defining aspect of competitive choice. This implies a diminution of loyalty. A customer who has bought business forms from one source for the past 20 years may suddenly switch suppliers and purchase based only on price as a result of the increasing margin erosion being felt by that business. Marketers who sell products with quantity-based pricing, like business forms, can influence and alter frequency by manipulating the price multiple. If a single form costs the business forms marketer $.02, then 1,000 forms cost $20.00. If the marketer works on a three-time mark-up, then the 1,000 forms will sell for $60.00. In order to obtain a $.02 manufacturing cost, the business forms marketer purchases in lots of, say, 150,000 forms at a time. If price was constant across quantity, the forms marketer would sell any number of forms for $.06 each to maintain the three-time mark-up. If a customer uses 60,000 forms a year, the marketer could sell those forms for $3,600 ($.06 x 60,000). The sale would be a one-time sale, a one-time shipment and a one-time billing. It is efficient. But if the customer wants to buy only a six-months supply of forms, or 30,000 quantity, the marketer has to make two sales efforts, two shipments and two invoicing cycles. The marketer’s costs of doing business increase. Therefore, the cost of the forms has to increase to cover the increase in business costs. As a result, the price for 30,000 forms is $.10 each, or $3,000 ($.10 x 30,000). If the customer chooses to conserve cash and purchases on a monthly basis, the unit purchase is 5,000 forms 12 times a year. The marketer’s costs of doing business with that customer go up substantially and, as a result, the price of the forms jumps to $.20 each or $1,000 for quantities of 5,000. Purchased monthly, the 60,000 forms used annually cost a total of $12,000. Purchased twice a year, the 60,000 forms cost $6,000. Purchased all at one time, the 60,000 forms cost $3,600. A frequency of 1 purchase a year delivers a multiple of three times cost; a frequency of 2 purchases a year delivers a multiple of five times cost; a frequency of 12 purchases a year delivers a multiple of ten times cost. Remember, the 60,000 forms, because they were purchased in lots of 150,000 quantity, still only cost the forms marketer $.02 each or $1,200 for 60,000 total. In other words, gross margin on a frequency of 1 is $2,400; gross margin on a frequency of 2 is $4,800; gross margin on a frequency of 12 is $10,800. Therefore, frequency drives quantity; quantity drives multiple, and multiple drives margin. If you were going to sell a total of 60,000 forms, who would you rather have as a customer? The 1-time, 2-time or 12-time frequency customer? And, in an increasingly hostile and cutthroat, competitive environment, which customer can you afford to offer a discount to keep the business? If you discount the 12-time frequency customer as much as 50%, the gross margin is $5,400, still higher than the 2-time frequency margin or the 1-time frequency margin. This is not an endorsement of discounting. But, from the quantity pricing examination of frequency, you can see why discounting is so prevalent in the commercial strategy of North America. The problems begin, however, when the discounting mentality begins to erode the gross margin levels even on the highest frequency purchases. Sad though it may be, it is possible that North American business will never see a whole price again. Discounting has been present ever since the first competitive transaction; it is a constant of our entire commercial history. But it has never been at the levels and the aggressiveness that is currently being practiced. There comes a point that you discount yourself out of business. In the final analysis, loyalty earned through discounting is a sham loyalty and must ultimately and utterly fail the discounter. Frequency and Product Satisfaction You buy a product. You love that product. You buy more of that or related products. Not a lot of magic there. Loving a product does not mean that you buy it at the same place every time. Product satisfaction and frequency do not always go hand in hand. As an example, you like the business forms you bought from Company A. The next time you need forms, six months later, Company B has similar forms on sale for 50% of what you paid at Company A. You buy from Company B. You are now a one-year recency customer and a 1-time frequency customer at both companies. Your definition of satisfaction is price and as long as you remain satisfied by price alone, Company A is never going to get a repeat order from you. Six months go by and you are completely out of forms. Without invoices, you cannot bill customers; in effect, you are out of business. Company B takes two weeks to fill your order, but Company A has a rush service and can deliver the forms to you the next day. This time, you buy from Company A. Your definition of satisfaction has changed from price to speed. Company B contacts you and says they are sad that you haven’t reordered. You tell them that they were too slow. They tell you to plan ahead and can offer you a program where you will receive forms every six months without fail and they can give you a VIP discount of 20%. You sign up for Company B’s program. Your definition of satisfaction goes back to price and now includes product usage management. Company A will never get an order from you again. Unless, of course, your definition of satisfaction changes based on circumstances. Satisfaction and frequency are moving targets. The customer can go along for five years, happy as a clam with Company B. And one day the telephone sales representative is rude and the entire satisfaction infrastructure collapses in five seconds. Hello . . . Company A? It is impossible to define satisfaction for every customer. It changes constantly and often irrationally. But existing within the customer base is an overall, qualitative definition of satisfaction provided that someone looks for it and is willing to devote massive amounts of time talking to customers, particularly customers with a one-time frequency. It is really so simple: Your job is to get prospects to buy and once the prospects have bought to keep them coming back; if they stop buying, your job is to find out why and to bring them back into the fold. More often than not, product satisfaction has less to do with low frequency and recency than with the way customers are treated, which, coupled with price, determines whether they come back. Price used to be secondary to service and, indeed, that may still be true. But, increasingly, price is primary and may even override service. If that becomes true, it is a sad day for North American commerce because it means that the customer has become resigned to lousy service and has decided to go for price since nothing else exists.
Frequency and Customer Service A happy customer is a frequent customer and a frequent customer is a happy customer. If price and quality are givens, then customer service, in the broadest, all-encompassing sense, is what keeps a satisfied customer coming back. For businesses selling consumable, commodity-type products, the most reliable indication of customer satisfaction is likely to be the frequency percentages. If a high percentage of customers are in the one-year recency, multiple-frequency category, the business is probably well run and the customers are probably well satisfied. If, however, a high percentage of customers are 1-time buyers and are spread equally throughout the one, two, three, and four years of recency, that is an indication that something is wrong. If product and price are not problems, then satisfaction is. It is possible to determine when customer satisfaction deteriorated or improved from frequency analysis. If a large group of high-frequency customers suddenly stopped buying two years ago, they will be seen as a "bubble" in the recency/frequency segmentation. At the point in time that the bubble formed, satisfaction dropped. Conversely, at the point in time that a bubble of one-time buyers suddenly became multiple-time buyers, satisfaction improved. By looking for these bubbles in frequency, it is fairly simple to isolate periods in a direct marketing company’s history when high out-of-stock situations existed, or when product quality was a problem or when customer treatment was less than perfect. If customers from the one-year frequency box who have a history of high- frequency purchasing over two, three or four years are isolated and examined, you will be looking at the best customers you have and, most likely, the customers with the highest customer service satisfaction levels. These are the heart and soul of the customer base and the people you must be talking to constantly. They can tell you what they like about doing business with your company, and that is some of the most important information you can ever obtain. Flag each of those customers, and when their frequency drops find out why immediately. The usual reason, price and quality being equal, is customer service satisfaction. That is the moment when the customer can be saved. Again, using frequency is common sense. If a customer walks into your store every three days and buys something and that customer has been doing that for the past three years, something is wrong when that customer doesn’t show up. That is when you have to find out what happened and what has to be done to get that customer back into a frequent buying habit. As a cataloger, you could have a million customers, but unless they are buying repeatedly they aren’t contributing profits to the company. It is better to have 1,000 customers buying four times a year for three years straight than to have 12,000 customers who buy only once. The total number of orders is the same, but the costs to get those orders are vastly different. The problem with customer satisfaction and frequency or recency is that rarely does a customer call and say, "This is my last order." They just get mad and stop buying and there is no known reason for their disappearance. As a result, maybe a year goes by and the customer is being mailed to every two weeks. At the end of the year, the customer is moved into the two-year recency box and is mailed to every month for another year. At the end of the second year, the customer is shifted to the three-year recency box and is mailed to every two months. It is conceivable that this customer could be mailed to for ten years and there isn’t a prayer of ever getting a second order. If you knew that up front, you could either fix the problem or cut out a lot of expensive mailings. But you don’t know that up front. You never know when you have just received the last order you will ever get from a particular customer. Using the database, recency and frequency information, it is possible to set up a proactive protocol that attacks the problem of potential customer loss. Simply stated, the database can be programmed to answer a series of questions. Q: Good morning, Database. As of this morning, how many one-year recency customers with a frequency of three or higher will move into the two-year recency category? A: Today, 214 customers with a 3+ frequency will age from one-year to two-year status. Q: Will you please rank them by highest frequency to lowest frequency? A: Certainly. Q: Will you please print out the names, addresses and telephone numbers of the top 20 customers on the list? A: I anticipated that request from your last inquiry. The list is ready. Here is where the work gets complicated. Contact each of those 20 formerly good customers and find out why they have not purchased in the past 365 days. This uses frequency analysis for a positive and beneficial reason. It is no different from the mental process the peddler used when going over the customer histories from previous trips. When somebody stops buying, there is always a reason. The only way to bring that person back into recency and frequency health is to know the reason and to solve the problem if one exists.
In the above example, the top 20 customers in the frequency hierarchy were isolated. In reality, the entire group of 214 customers for that one day would be ranked by frequency and acted upon as appropriate. But there do exist parameters. The hierarchical ranking of 214 customers represents a value ranking. Presumably the customer at the top has the highest value relative to the frequency of purchases. The customers at the bottom — the one-time frequency customers — have a lesser value. All are important, but if you were going to actively work to get one back it would be the customer that has bought from you 12 times in a one-year period (provided the average order size is equal for all customers. More on this in the chapter on monetary value). The same hierarchy of value exists relative to frequency and customer mailings. If all two-year recency customers receive 8 mailings a year, perhaps 12 mailings to two-year recency customers with a 6+ frequency is not such a bad idea. Perhaps ten mailings to two-year recency customers with a 3 to 5 frequency is a better strategy than just the uniform eight mailings that "all two-year customers" would normally receive. Logic would dictate that a one-year customer with a four-time frequency spaced at three-month intervals is probably a candidate for no more than four mailings a year spaced at three-month intervals. Too often direct marketers attempt to make the purchase fit the number and timing of the mailings rather than making the mailings fit the customer’s frequency purchasing history and schedule. If you can obtain the same sales level with four mailings versus eight mailings, you are a direct marketing rainmaker. Obviously, these assumptions must be carefully conceived, tested and tracked for profitability over a sufficiently large sample and an adequate length of time or number of offers. But recognizing that frequency analysis offers many other possibilities for the direct marketer’s arsenal of strategies besides just "who buys most often" is important.
If the 80-20 rule is alive and well (and it is), there will be a correlation between customer frequency and product propensity. If 20% of the customers are delivering 80% of the sales, then it is likely that 20% of the products are delivering 80% of the sales. If you know which customers and what products, you are well on your way to the Big Money Machine. The ideal marketing database allows "give me" questions such as, "Give me all of the one-year customers with 3+ frequency who have bought four or more products in each purchase; then give me a hierarchy of frequency and products with the most often purchased products listed in descending order." In the same manner, you can diagnose frequency problems by asking, "Give me all of the two-year customers with a 3+ frequency who have not purchased in the past 12 months; then give me a hierarchy of frequency and products with the most often purchased products listed in descending order; then give me a hierarchy of products with the highest returns or replacement rates for that same period of time and correlate that hierarchy to those customers. Is there a valid correlation between product dissatisfaction and frequency drop-off?" By going inside the shoebox and peering into the smaller shoeboxes of recency and looking at the compartments of frequency in each one, specific questions can be asked and answered about the dynamics of products, pricing, delivery speed, shipping and handling charges, advertising changes and a host of other questions important to the foundation objective of any direct marketer: selling more of the products that the best customers want.
A bubble of frequency drop-off in the month following an increase in prices may tell you something. Similarly, a bubble of frequency drop-off following an increase in shipping and handling charges may be saying something about what the aggregate customers think about that added expense. Products that are reduced in price or offered as a loss leader may drive bubbles of increased frequency for the life of the offer. The measure of the value of such offers is whether the frequency of purchases remains higher after the offer than before the offer. If a customer can be converted on price and maintained at a satisfactory frequency level thereafter, then the price strategy was successful. But if the loss leader only produces a momentary blip in frequency and the customers return to their former lethargy, then nothing has been accomplished except margin erosion.
Direct marketing companies that sell products in quantity generally price the products attractively in the higher quantity levels. As discussed previously, the multiple for higher quantities is less than that for smaller quantities. There may exist a four-time mark-up at 100 quantity and only a two-time mark-up at 1,000 quantity. Ideally, if every customer purchased at a higher multiple and a lower quantity frequently, the business would operate at a significantly higher gross margin. Some direct marketing companies make decisions on quantity price breaks and the merchandising of potential savings without fully understanding what happens to frequency and, ultimately, profitability. As a result, advertising suddenly extols the virtue of buying larger quantities less often. Words like "savings" and "convenience" and "never run out" appear in the advertising. Big arrows are pointed at the price for 1,000 accompanied by a bright yellow sunburst screaming "as little as two cents each in quantity." The customers recognize the savings and actually increase their quantity- level purchases to take advantage of the price break. Of course, they are buying the same annual quantity in one purchase instead of three purchases. Suddenly, frequency drops from an average of three times a year to once a year. Margin drops sharply, the bottom falls out of profitability and the company asks, " Why?" Experienced direct marketers understand the effects of decreasing frequency and margin while maintaining units. It is a prescription for disaster. But inexperienced direct marketers see it as a strategy for increasing sales dollars, which it can do up front. The inevitable outcome, however, is that increased sales at lower margins without corresponding increases in total units sold is simply a form of accelerated margin erosion leading to financial collapse. The only benefit that results is the vacant position in the marketing department. Frequency and Advertising Space
Frequency of product purchases can be influenced by the amount of advertising space devoted to the merchandising of products. If, on average, customers purchase blue checks 2.6 times a year when blue checks are displayed on a half-page of advertising, what happens when blue checks are squeezed into a one-quarter-page advertisement? Maybe nothing, or maybe a lot. If price, quality and satisfaction are constant, and demand measured by frequency drops, then the product and frequency are influenced by the amount of advertising space. Possibly the old maxim "out of sight, out of mind" applies. If a customer always bought two yards of blue cloth from the peddler and on one trip the blue cloth was tucked away in a corner of the wagon instead of being in full view, it is possible that the customer would not see the blue cloth and, as a result, not purchase it on that trip. Frequency is at the heart of the shelf space wars in grocery stores. The more prominent the product, the more likely the customer will buy it — often. The same common sense principle applies in direct marketing. If you want a customer to purchase a product repeatedly, then that product has to be displayed prominently and constantly merchandised. If one product accounts for 38% of sales, it is logical to conclude that a lot of space should be devoted to selling that product. Too often, frequency analysis is not linked to product propensity and, subsequently, to advertising space allocation. And yet, perversely, every direct marketer recognizes the value of that logic. Getting it done is the problem.
Frequency velocity is the speed at which the number of purchases is increasing for a particular customer and, as an aggregate, for the entire customer base over time. As an example, if the average customer moves from one purchase a year to four purchases a year, the frequency velocity goes from 1 to 4, a net increase of three purchases (+3). Conversely, if the average customer goes from four purchases a year to one purchase a year, the frequency velocity is then a net decrease of three purchases (-3). The assumption is that a positive increase in frequency velocity is also a positive increase in total quantities and dollar value of products bought. A customer for business forms or other commodity products who simply increases the number of purchases without increasing the total number of units bought is only benefiting the direct marketing company through the increased margin on the multiple for the smaller quantity of product bought each time. While that is an improvement, it pales alongside the customer who buys more product and other products thereby increasing the overall average order value and lifetime net profit. Dividing the customers into descending hierarchies of positive frequency velocity and negative frequency velocity can provide potentially important knowledge. By studying all customers with, say, a +3 frequency velocity relative to products purchased may reveal product propensity information that can be used to stimulate similar increases in purchases from similar customers at the +1, +2, or static levels of frequency velocity. In the same manner of study, it may be highly desirable to isolate product commonalities among customers who move from a positive frequency velocity to a negative frequency velocity. Equally important are studies of customer frequency velocity — either positive or negative — relative to changes in pricing, shipping and handling charges, fulfillment delivery time, number of mailings, reorder reminder initiatives, package inserts, sale or discount offers, rush charges, coupons, premiums, product maturity and the host of other novelties and nostrums frequently used in fully furnished direct marketing programs. This is the stuff that moves the needle. A positive increase in overall customer frequency velocity of 1/2 of one order annually is a net increase in total orders of 50%. Knowing that a decision or an action either moves the purchasing velocity upward or downward is a very important bit of knowledge. Too often, such decisions and actions are put in place with no measurement whatsoever of the effect on sales other than, "Gosh, it looks like we goofed." Knowing early that 5% of the +3 frequency velocity customers have dropped to a +1 velocity is an important bit of strategic information. The velocity is an indication of the heartbeat of the business. If the heart is slowing and pumping less blood, knowing that early is far better than knowing it after the business has slipped into cardiac failure. These are subtle measurements. They are apparent only upon close examination over time. You must have standard reference points from which to plot the subtle movements and the patience to develop the measurement system and to let it percolate until it becomes meaningful. Again, keep the conceptual thinking simple. Go back to the peddler and the shoebox. If half of the customers on the route suddenly stop buying on one of the two annual trips, the business has just diminished by 50% (as long as the average order value remains constant). Clearly, something has happened and you have to find out what it is. So you take the top frequency customers who had been adding purchases at the fastest rate and ask them, "Why have you decreased the number of purchases?" It is simply amazing what the customers will tell you if you ask the question.
Frequency momentum is the aggregate force of the frequency velocity in the customer base. If the customers are, for the most part, all adding purchases and increasing their individual frequency velocity, then the customer base has a positive frequency momentum. On the other hand, if the majority of customers are decreasing their individual frequency velocity, then the aggregate measurement of momentum is negative. If you have 100 customers buying one time a year in each of three years in a row, the frequency velocity is zero. It is neither positive nor negative. If ten customers increase their purchases to two a year, the aggregate frequency velocity is +0.1. If 50 customers increase to two purchases a year, the aggregate frequency velocity is +0.5. If all 100 customers increase from one to two purchases a year, the aggregate frequency velocity is +1. The rate of frequency momentum — the measurement of the force of frequency velocity — would be 10% for the +0.1 velocity above. It would be 50% for the +0.5 velocity and 100% for the +1 velocity. In other words, aggregate customer frequency momentum is increasing by 10%, 50% and 100%, a good thing. Of course, velocity decreases describe an opposite and negative frequency momentum, not a good thing. The value of these two measurements of frequency movement is more apparent when considered over time. For example, a business may have a rapidly improving frequency velocity but the overall frequency momentum has not turned from negative to positive. Turning a business is like turning an ocean liner. The speed of the ship is the velocity; the distance it takes to turn the ship is the momentum. You may be increasing the speed, but the ship has yet to make the turn and another three nautical miles have to be covered at an increasing speed before it is back on course. Distance is time; time is money. Positive frequency velocity and momentum are money.
Thus far, recency and frequency have been examined as analyses of a direct marketing company’s health or strength. The focus is always on improving customer retention and increasing the average order value. But both recency and frequency have a great deal to offer in the process of evaluating a direct marketing business for acquisition. Generally, two types of direct marketing companies exist relative to acquisitions: 1) successful businesses that sell at a premium and 2) distressed businesses that sell at a discount. The first step, for the acquiring organization, is to determine which of these two conditions exists. For an experienced direct marketer, this is not too difficult; for an inexperienced buyer, this determination can be fraught with unknowns. Ideally, an acquisition worthy of a premium purchase price will have positive recency velocity, positive recency momentum, positive frequency velocity, and positive frequency momentum. The vital signs of such a business all point to health, vigor and vitality. An acquisition worthy of only a discounted purchase price will have one or more of these key indicators trending negatively. Commonly, profitability is the cardinal indicator of acquisition price. Unfortunately, profitability does not go far enough towards isolating the reason for profitability. Almost any direct marketing business can be shown to be profitable if all prospect mailing for new customers is suspended. If 50% of the advertising budget is spent on new customer acquisition mailings, that money suddenly falls to the bottom line when those mailings are stopped. But the recency and frequency velocity and momentum are deteriorating and, in a short period of time, the business will be in distress without new customers being added daily. Given an acceptable average order value, the true worth of a business is in the proven customer performance relative to recent purchasing and frequent purchasing. If those two vital signs are moving positively, the business is healthy; if not, the business is diseased. Unless distressed business turnarounds are your specialty (often a lucrative strategy), the logical acquisition strategy is to purchase a company that is successful. The difficulty in evaluating direct marketing companies for acquisition comes in assembling the necessary data. Good companies are aware of recency and frequency analyses and can demonstrate their performance; bad companies most likely wound up in trouble because they don’t have any performance information. But knowing what you need to know in order to make an informed decision is the first objective, and knowing recency, frequency and monetary value is the first step in achieving that objective.
All frequency reconciliation takes place in the one-year recency box as all purchases, whether first-time or repeat, are accounted for there. Frequency reconciliation can be accomplished on a daily, weekly, monthly, quarterly and annual basis. If you choose to do daily reconciliation, you will know more about frequency performance faster. If you choose to do annual reconciliation, you will minimize the available knowledge. In a perfect analytic world, daily reconciliation leading to weekly, monthly, quarterly, annual and a 20-quarter rolling reconciliation would provide adequate frequency information, understanding and knowledge. Three numbers are required for accurate frequency reconciliation: 1) total number of orders, 2) total number of first-time or new orders, and 3) total number of repeat orders. If 1,000 orders are received today and 200 of those orders are first-time orders, then 800 orders are repeat orders. Clearly, the first requirement for accurate frequency reconciliation is the ability to separate first-time orders from repeat orders. A method to categorize orders from first-time customers and orders from repeat customers is essential. Once the 200 first-time customer orders have been assigned to the first-year recency box, the 800 repeat orders can be classified. Of the 800 orders, some number will come from the one-year recency segment, some from the two-year recency segment, and some from each of the three-year, four-year, and five-year- and- older segments. If 400 orders are from the one-year box, 200 from the two-year box, 100 from the three-year box, 75 from the four-year box, and 25 from the five-year and older box, then the entire 800 repeat orders are accounted for and the reconciliation is accurate. From that knowledge, an evaluation of the performance in each of the recency segments relative to advertising efforts directed to each segment can be done. Remember, for frequency reconciliation, you are counting orders; for recency reconciliation, you are counting customers. With 20 quarters of accurate frequency reconciliation data, it is possible to spot trends in the customer reorder patterns. For example, knowing that you are getting a higher number of repeat orders from the four-year recency segment than the two-year recency segment is indicative of something going on that you should be aware of. It may be that the two-year recency segment was inadvertently left out of a mailing, or it may be that old customers are returning because of price changes by your competition. Many possibilities exist for variances in frequency performance; knowing who, what, where, when, why and how much through solid analysis and customer contact is the difference between survival and extinction. In addition to the frequency/recency blend of analysis and knowledge, the frequency reconciliation is the only information that will tell you whether the frequency of individual and aggregate repeat customer purchases is increasing or decreasing. In short, unless you reconcile accurately, you will have no indication of frequency velocity or momentum either for the individual customer or for the customer base as a whole.
As with recency, a language exists to describe frequency. Customers are classified in one of two major segments: 1) one-time buyers or 2) multiple buyers. A one-time frequency buyer has a history of only the initial purchase; a multiple buyer has a history of repeat purchases. The value of the multiple buyer is apparent and a premiu2m is placed on these customers. In not-for-profit or fundraising, these frequency terms are one-time donor and multiple donor, or similar descriptive terms. A one-time buyer can be found anywhere in the recency progression. Beginning with the first year, the buyer can age through the second, third, fourth, and fifth years of recency without ever becoming a multiple buyer. A multiple buyer can also be found in any of the recency boxes, preferably in the one-year file. However, multiple purchases can be classified in two ways: 1) multiple purchases in the aggregate or 2) multiple purchases in the one-year recency segment. A customer who purchases once a year for five years in a row is a one-year recency customer with a multiple purchase history of five separate purchases. A customer who purchases five times a year, however, is a one-year recency customer with multiple purchases in the same year. The first customer has one purchase and the second has five purchases in the same period of time. Both customers are multiple buyers and both customers are one-year recency customers, but they are very different. If both customers stop buying and age through the recency categories, they will age as multiple buyers but of different strengths. The old five-times-a- year buyer is still more attractive to the marketer than the once-a-year buyer (provided that the average order value is the same). So that these differing strengths of frequency can be used to advantage, it is necessary to segment the customers in the database logically. At the minimum, this calls for three frequency segmentations: 1) one purchase only; 2) one or more purchases in the aggregate, and 3) 2+ purchases in the one-year recency box. The first describes all customers who have bought only once. The second describes all customers who have bought more than once during their life as a customer. The third describes all customers who have made two or more purchases during the current year. Clearly, a hierarchy of frequency value exists in these three classifications. From there, the second and third segmentations can be further subdivided into actual numbers of purchases. For example, customers who have made 2 purchases in the aggregate, 3 purchases in the aggregate, or 12 purchases in the aggregate. In the same manner, the one-year frequency segmentations could be customers who have made 2 purchases this year, 3 purchases this year, or 12 purchases this year. Unless purchases are accounted for individually as fulfillment of an obligation, such as a continuity program like a book-of-the-month club, there may be little value in detailing individual numbers of purchases. Generally, frequency is subdivided into blocks of purchases, such as 1, 2-4, 5-8, 9-12 and 12+. The blocks are entirely dependent upon the needs of the company for analysis and the flexibility of the database system software. The important ability is the differentiation between the one-time, multiple aggregate, and multiple one-year buyers. The continuity of the analyses must be assured. You cannot change analysis categories down the road and expect to get oranges and oranges. Starting with a greater detail of frequency segments may not be a bad idea. You can always combine them, but you may not always be able to separate them. Nothing is more frustrating than to have information but not be able to attach accurate meaning to it because the yardsticks were changed.
In Chapter I, the relationship between recency and list rentals was explored. Many of the list rental considerations for recency apply to frequency as well. If a one-year recency customer is a better customer than a three-year recency customer, then a one-year multiple buyer is a better customer than a three-year, one-time buyer. When recency and frequency are combined, gradations of list rental quality and attractiveness become clear. A five-year-and-older, 1-time buyer would have the least value as a rental; a one-year, 12-time buyer would have the greatest value. And between these two recency/frequency extremes are the gradations of quality and efficacy that determine the value of the segmented list. A list renter may specify one-year customers with a frequency of four or more purchases. Because the quality of these names is greater, so is the rental charge. Of course, as the frequency level increases, the total number of names decreases. At a frequency of 12+ purchases in the current year, a very small number of customers is likely to be found. In order to obtain the necessary volume of names, a renter may work downward in frequency or may simply specify all 2+ buyers in the one-year file from the start. Generally, a premium is charged for renting multiple buyer names over one-time buyer names. The justification for the higher rental fee is the proven quality of the customer. The quality is worthy of testing. A four-time frequency, three-year recency customer may not necessarily have stopped buying through direct mail. That customer may have simply stopped buying from that particular company. By renting old customers with high frequency, it is possible to test the quality and, therefore, the potential value of the list at a reduced rental cost. Remember, this is a business of pennies. One penny over one thousand names is $10; over one million names it’s $10,000. Again, at the minimum, the three segments of recency (one-time buyers; 2+ aggregate buyers; 2+ one-year buyers) should be explored in list rental strategies. As a company renting its list, there can be incremental income value gained through offering enhanced segmentations of combined recency and frequency. Few direct marketing companies have established a "menu" of rental charges that begins with, say, premium-priced 5+ one-year multiple buyers and ends with bargain-priced one-time, five-year-and-older buyers with all of the various segments in between value-priced in scale proportionately. Frequency and Offer Life and Half-Life
Knowing how long it takes a customer to make a decision to buy, stimulated by advertising, is essential to forecasting order flow and cash flow. As an example, if all first-time buyers order within 60 days of the receipt of advertising, the life of an offer can be said to be 60 days in length. If 50% of the orders from first-time buyers are received by the fifteenth day, the half-life of the order is 15 days, the half-life being the point at which 50% of all orders are received. The balance of the orders will arrive over the following 45 days. But what is the life of an offer and the offer half-life for one-year multiple buyers? Because they are already customers with experience, it may be that the life of an offer is shorter and the half-life faster. All orders may be received from this customer segment within 30 days and the half-life may be at 10 days. By shortening the half-life from 15 days to 10 days, cash flow is improved. Staffing patterns also are predicated on order life and half-life. As frequency velocity and momentum improve, order life and half-life should also improve. Staffing a direct marketing business for a lengthened or shortened half-life and order life is important strategic information. The mix and timing of prospect mailings versus customer mailings can influence these factors significantly. The life of an offer in days and the half-life of an offer in days can be used to validate the measurements of frequency velocity and momentum. If the half-life is extending out, it may be an indication that frequency momentum is falling. The customers who normally buy soon after the offer is received may not be performing to past standards. Knowing this is essential. Not knowing this is suicidal. The life and half-life of an offer is crucial to maintaining efficient inventory levels. If the cycles shorten, out-of-stocks can occur; if the cycles lengthen, too much money can be tied up in inventory that is not turning. Precision forecasting demands awareness of the order life and half-life measured in days and the dynamic changes taking place within those measurements. This is sophisticated stuff. Not too often do you run across a good chart of a 20-quarter rolling average of 2+ one-year multiple buyer half-life and order life compared to a 20-quarter rolling average of first-time buyer half-life and order life. Even less often do you find direct marketing companies that actually stock inventory and schedule employees as a result of this knowledge. The direct marketing CEO who works through this infinitely detailed information regularly, looking for minute signs of changes and emerging trends, is the direct marketing CEO that will remain competitive and cost-effective in an increasingly hostile economic and social environment.
No analytic tool is worth anything unless it can be used to create and keep customers. Information without sales is useless, bordering on commercial auto-eroticism. Multiple buyers, be they business or consumer, have buying patterns. The business check buyer often purchases on a quarterly basis with an additional purchase or larger purchase at the end of the calendar year. The consumer check buyer tends to purchase 1.98 times a year, a pattern driven by the fact that 33 checks on average are written each month for an annual usage of 396 checks and check orders are generally for a 200 quantity. The ice-melting chemical customer purchases in the Fall of the year and may reorder after the first of the year if the season is particularly rough. Knowing these patterns of frequency are incredibly important if the marketer is to intervene when an expected purchase does not occur when it should. Frequency is the first indication that something has gone wrong with a customer. When a lapsed customer shows up in recency, it may be too late by one or more orders. Frequency tracking and the expectation of a customer’s frequency pattern are the early warning devices necessary for an optimal retention marketing strategy. So often, direct marketers overly complicate frequency pattern expectation. The method for determining when a customer may have lapsed becomes so cumbersome that it fails to work. In reality, using frequency pattern expectation is no different from using a 3 x 5 card tickler file. If a customer buys every 90 days, the tickler file card is advanced 90 days and checked. If the customer bought, fine; if not, find out why. The peddler used this same system with great success. If you think about it, frequency pattern expectation is no different from a reorder reminder system. Every 90 days, or whenever another purchase is expected, a reorder reminder is cranked out of the database and the customer is asked for another order. When the pattern changes, the reorder reminder pattern changes. And all too often that is where the efforts stop. The successful direct marketer is not satisfied with sending a reorder reminder and waiting for something to happen. If the order does not arrive when expected, the successful marketer is on the phone finding out why. That is the point where customers are saved, where customers are retained, where the enormous cost of obtaining that customer is justified. In its pure and simple form, frequency pattern expectation should produce three lists every single business day: 1) customers who should order today, 2) customers who did order today, and 3) customers who did not order today. Any customer appearing on the lapsed customer list is a candidate for immediate action. Frequency pattern expectation requires some buying history, especially when the product is not a consumable product. But even for first-time buyers, some generalization can be made as to when another order should be expected. Start there and adapt to the individual customer’s pattern as it evolves over that customer’s purchasing lifetime. Adaptive frequency pattern expectation can be an exceptionally dynamic marketing technique. By adapting to the demonstrated timing of the customer’s changing frequency pattern, the reorder program learns. For example, consider the business check buyer who orders 1,000 checks quarterly without fail for a three-year period. The reorder reminder is sent 70 days after the last order was received, allowing a 20-day reorder window before the customer runs out of checks. Suddenly, the customer re-orders for 30 days instead of the expected 90 days. The quantity of checks is smaller and equals a 30-day supply. At this point, an adaptive frequency pattern expectation program would conclude that the customer has switched to a monthly reorder pattern for some reason. Two adaptive actions would then occur. First, the customer should be called to verify this change in pattern and, second, the reorder reminder schedule for that customer would be moved to a monthly pattern. If, later, the quarterly reorder pattern is returned to by the customer, the reorder reminder program readapts to that frequency and timing change. The reason for the change in pattern may be one of any number of factors, perhaps cash flow; the marketing response is less concerned with the reason than with meeting the customer’s changing frequency pattern need. Frequency and Why It’s Not Quite Enough Frequency when combined with recency is far more revealing and effective than recency alone. But it is still not enough because it doesn’t quantify the dollars spent. To illustrate the progression in the useful value of recency, frequency and monetary value (and to set up the next chapter on monetary value) consider the following comparative case. You own a paint store. Your customers buy paint and all the products needed to apply or remove paint. For the first ten years in business, you track customer performance only by recency. If a customer does not show up every year and buy something, you send a postcard asking for that customer’s business. That works pretty well, but you have a large number of customers who buy only once and never come back. You suspect that price has a lot to do with retention and decide to concentrate on getting more sales from your existing customers. So you begin tracking frequency. For the next ten years, you use recency and frequency and are able to isolate the 20% of the customers who make 80% of the purchases. You work hard to keep the customers coming back. Only you are not making a lot of money. One day you take out the shoebox and go over the customer cards. You find out that you have, basically, three types of customers: 1) one-time-only buyers; 2) one-time, one-year recency buyers; and 3) one-year recency, multiple buyers. The one-time-only buyers bought once and haven’t been back for several years. The one-year recency, one-time buyers come in once a year and buy something. The one-year recency, multiple buyers come in about monthly and buy something. You decide to concentrate on the one-year recency, multiple buyers and spend a lot of money on advertising to them with special offers and VIP customer plans and all sorts of good stuff. Your rationale is logical: A multiple buyer is a better customer than an annual customer, and an annual customer is a better customer than a one-time customer. You have divided the customers into three segments of value. But here is the reality. The multiple buyer spends on average $5 on each of 12 purchases a year for a total of $60 annually. These are small, fix-up paint customers. The annual buyer spends $140 every year on one purchase. These are the regular maintenance paint customers. But the one-time buyer spends $1,700 every five years. These are the customers who own horse farms and have to paint six miles of white fences every five years. Now which customer are you going to wine and dine? The multiple buyer spends a total of $300 over five years and requires 60 individual sales. The annual buyer spends a total of $700 over five years and requires five individual sales. The one-time buyer spends $1,700 over five years and requires one sale. And that is why recency and frequency are not quite enough. Real-Life Monetary Value Characteristics Monetary value is how much a customer spends. It is an indicator of dollar value. As an aggregate of all customers, it is a primary measurement of average order value and determines total sales volume. By itself, it is arguably the most forceful indicator of the cardinal three — recency, frequency and monetary value (RFM). Monetary value is where the rubber meets the road. You see, unlike the fairly tame concepts of recency and frequency, now we are talking about money. Monetary value describes a state of a customer’s worth. If you think about it, there are only two states that describe customers: profitable and unprofitable. Profitability is relative to the amount spent versus the cost of the product, services provided, and overhead. A customer either spends enough to be profitable or does not. Monetary value measures that spending. From the paint store example at the end of the last chapter, it is evident that totally wrong conclusions can be drawn from recency and frequency analysis unless they are made in combination with monetary value. Only monetary value can validate the apparent hierarchy of customer value described by recency and frequency. To rely on recency and frequency without monetary value as the measurement of direct marketing performance is dangerous. The best analogy is your checkbook. It is good to know how many checks you have written and to whom they have been written, but unless you know the amount of each check, you are flirting with economic self-destruction. As a rule, monetary value (expressed for this part of the discussion as average order value, or AOV) rises in tandem with inflation, assuming a business is recovering its inflationary cost increases through regular product and service price increases. If inflation has been 5% in each of the past three years, then monetary value is likely to have increased by 15.76% in price increases over the same period of time just to remain even. If, however, monetary value has increased by 30% in the same period, and if the price increases were limited to the 15.76% inflation factor, then real monetary value growth has been 14.24%, a strong indication that customers are purchasing more. Determining real monetary value increase or decrease is essential to understanding what this measurement of direct marketing performance means. Monetary value, adjusted for inflation, can increase, decrease, or remain static; no other choices exist. Only an increase in monetary value over time is a symptom of robust business health. Static or declining monetary value over time are symptoms of disease and must be treated. If monetary value is static, the business is experiencing margin erosion as a result of natural inflation. If monetary value is declining, the business is experiencing margin erosion due to both inflation and decreasing customer purchases. The first is critical; the second is terminal. Monetary value relative to total sales must also be considered. If monetary value is increasing and total sales are increasing, the symptoms are positive. If, however, monetary value is declining and total sales are declining, the symptoms are grave. The nuances, where monetary value is static and total sales are declining, or where monetary value is declining and total sales are static, are indications of other than positive influences on the business. Again, as in recency and frequency, the historical record and trend over time must be illustrated and understood. A rolling 20 quarters of monetary value history compared to total sales is essential to this understanding. To cover the basics, consider the following example. Monetary value (average order value) is $1.00. Total sales are $1,000. Therefore, the total number of orders is 1,000. If customers order on average twice a year, the customer count is 500. If the customer count remains at 500 and the number of orders remains at 1,000, but the total sales increase to $2,000, then the AOV must have increased to $2 per order, an indication that customers are either buying more or that prices have doubled. Of course, it is also possible that one customer ordered $1,000 worth of product. Diabolical, isn’t it? What the above example stresses is the importance of knowing exactly what is going on with the money customers are spending. That is what monetary value is all about. Monetary Value and Segmentation Return with us now to those thrilling days of yesteryear when shoeboxes and 3 x 5 cards were the mainstay of direct marketing analyses. All of the cards are ordered by recency and frequency. You can pick out the most recent buyers and the most frequent, multiple buyers. Wouldn’t it be nice if you could identify the cream of those customers who spend the most money? First, determine what dollar levels of spending are realistic segments of monetary value to track for your company. 2Any range of spending can be established, but must be useful, logical, and constant. Changing the monetary value categories down the road can invalidate the information that develops. Begin with your average order value and work upwards and downwards in value. For example, if the AOV is $150, you will most likely be interested in $25 or $50 segments of value. You could establish the following segments: $0-$25, $26-$50, $51-$75, $76-$100, $101-$125, $126-$150, $151-$175, $176-$200, $201-$225, $226-$250, $251-$275, $276-$300, and $300 plus. By distributing every order into its category of dollar value, you will segment all orders by monetary value. The greatest number of orders will fall into the segments on either side of the average order value; lesser numbers of orders will fall into the segments at the extremes. In effect, a bell Curve will form with the bulk of orders at the average order value. Because an order is a customer, you now have the ability to say, "Give me all of the orders and the customers between $251 and $275 in value." That listing is a unique segment of the customer base and tells you something unique about those customers: They spend $100-$125 more than the average customer. Most likely, you will rank and list all customers in a descending hierarchy based on monetary value. At the top will be the one customer who has spent the greatest amount of money; at the bottom, the one customer who has spent the least amount of money; in between, all the rest. Once you have the hierarchical ranking by small segments of monetary value, it is usual to lump segments together and look at overall monetary value performance groups. Generally, these groups can be described and classified as either being 1) unprofitable 2) profitable or 3) very profitable. For example, if the AOV is $150, set up three monetary value groups at $0-$100, $101-$200, and $201-$300 plus. If you must have a minimum order of $101 to cover fixed, overhead and advertising costs in order to make a profit at your product mark-up, the first monetary value group at $0-$100 is unprofitable; the second group at $101-$200 is profitable; the third group at $$201-$300 is very profitable. Some marketers refer to these three classic groups as Dogs, Cows, and Stars. Logically, you want to market aggressively to Stars, steadily to Cows, and carefully to Dogs. The intent is to raise Dogs to Cows, Cows to Stars, and Stars to Superstars while attracting no new Dogs. Do that and you will be a Superstar. Caution must be urged at this point for there are many ways by which monetary value can be segmented and measured. Assuring continuity from measurement to measurement, from hierarchical listing to hierarchical listing, from analysis to analysis, and from comparison to comparison is crucial if the knowledge is to have any validity. One way to construct segmentation by monetary value and to generate the hierarchy is to look only at the last or most recent purchase across the entire customer base. Whether that purchase was this morning or five years ago, only one purchase per customer will be included in the hierarchy and ranking. A second way to construct segmentation by monetary value and to generate an entirely different hierarchy is to look at the aggregate of all purchases for all customers across the entire customer base. This is the lifetime aggregate monetary value and it provides knowledge of a different type. A third way to segment and rank is by year of recency. For instance, a hierarchical listing of all segments of monetary value in only the one-year recency classification can be done. Similarly, a monetary value ranking for all two- or three- or four-year recency customers could be accomplished for comparison. Obviously, each of these recency segments could be analyzed for either last purchase only or aggregate lifetime purchase monetary value (and each would be different). Before the complexity of monetary value becomes overwhelming, remember that there are only a few questions that are really important: 1. Is the average order value increasing relative to sales? 2. Is the ratio of Stars to Cows improving? 3. Is the ratio of Cows to Dogs improving? 4. Is the monetary value performance in each year of recency as good, better than, or worse than last year (or some other comparative time period)? 5. Is the monetary value for most recent orders relative to the monetary value for aggregate lifetime orders, or is there a surprise in the making? If you know the precise answers to those five questions, you will be ahead of 90% of the competition; maybe 99%. As with recency and frequency, monetary value velocity is the speed at which the individual and aggregate customers are increasing their dollar purchase levels over time. As an example, if the average customer moves from an average order value of $150 to an average order value of $200 in one year, the monetary value velocity goes from 150 to 200, a net increase of 50. Conversely, if the average customer goes from $150 to $100, the monetary value velocity is a net decrease of 50 (-50). Any increase in monetary value velocity is positive; however, the greater the velocity, the greater the dollar level of spending. Dividing the customers into descending hierarchies of positive and negative monetary value velocity can provide potentially important information. By studying all customers with, say, a +100 monetary value velocity relative to products purchased may reveal product propensity information that can be used to stimulate similar increases in spending from similar customers at the +25, +50 or static levels of monetary value velocity. In the same way, it may be very instructive to isolate product commonalities among customers who move from a positive monetary value velocity to a negative monetary value velocity. Individual monetary value velocity represents customers who suddenly are spending more money than they have historically spent over some period of time. Something has motivated these customers to purchase more; knowing what that motivation is, when it occurred, and how much it creates in additional monetary value are important pieces of information. Equally important is the ability to use monetary value velocity to measure the effect of pricing, shipping and handling cost increases, product mix alterations, shifts in marketing emphasis, mailing frequency and other changes to your business. As an example, if you raise prices 10% across all products and the monetary value velocity on an average order value of $150 only increases by $7.50, you have a 7.50 decrease in velocity. This represents a 5% decrease in adjusted average order value, most likely as a result of price resistance. In the above examples, monetary value velocity has been expressed simply as a number equaling the dollar difference. It can be calculated any number of ways — as a percentage, as a dollar amount, as a score — as long as the calculation is consistent over time. The end product of the measurement is to be able to state, "This customer has improved by (some appropriate measurement)," or "This customer has deteriorated by (some appropriate measurement), and it has occurred over the past (some period of time)." That is the kernel of knowledge leading to wisdom that justifies doing these analyses. Once a clear understanding is gained about the health of the monetary value velocity, the information can be examined relative to recency velocity, frequency velocity, recency momentum, and frequency momentum. From those examinations, precise understandings of pockets of customer behavior will be gained and optimal customer segments will surface. Monetary value momentum is the power or the force being exerted on the customer base as a whole as a result of changes in monetary value velocity. If the velocity is increasing at a faster and faster rate, the momentum is positive. If the velocity is decreasing at an accelerating rate, the momentum is negative. For example, if, over the last three mailings, the $150 average order value increased by $4, $6, and $8 successively, the average order value stands at $168, an overall velocity of 18 and a momentum of 6 (18 divided by 3). If the AOV increases by $12 on the fourth mailing, the velocity will be 30 and the momentum 7.5 (30 divided by 4). The momentum is increasing by 1.5. It is possible to have a positive, slowing momentum and also a negative, increasing momentum. These nuances describe the trend of the momentum — up but starting to slow; down but slowing or improving faster; or staying the same. Knowledge of the trend allows you to say with certainty, "We are recovering our recent, six-month-long loss of position in monetary value at an accelerating rate." In order to make the smart decisions, you have to know exactly where you are and in what direction the momentum is flowing. With the addition of the monetary value velocity and momentum component, it should be clear now that the best customer you can possibly have is the customer with an accelerating momentum and an increasing velocity in recency, frequency and monetary value. If the trends in each component area are up and gaining speed, the customer is a top-of-the-line buyer and one that must be focused on with obsessive concentration. Reconciliation of monetary value is a process of counting customers rather than orders. Each customer has a monetary value classification, determined from either the last purchase or the aggregate of all purchases. The classifications are the segments of dollar value that you have chosen for your business. The segments can be in small dollar increments, such as $50-$75, or in larger dollar groups, such as Stars, Cows and Dogs as discussed previously. At the end of any business day, the shuffling and arranging of individual customers in the monetary value segments will change based on that day’s purchases. A certain number of customers will be in the highest monetary value classification; a certain number in the lowest; and the balance spread throughout the classifications in between. The new customers will be assigned a monetary value classification based on their first purchase spending level; repeat customers will either remain in their existing monetary value classification or move up or down depending on their purchases. But every customer can be accounted for and located into a monetary value box. Tomorrow that reconciliation will change. Customers will move to other monetary value classifications. New customers will be classified for the first time. The total number in each box will be different from the total number today or yesterday. Movement of customer monetary value — like movement of customer recency — is a dynamic phenomenon. Monetary value classifications — measured in dollars — will each have a total number of customers "residing" in that classification. The changing totals over time are the fruit of the analyses. Are the customers flowing in the direction of increased monetary value or are they flowing toward a diminishing monetary value? Obviously the average of all customer monetary value is the average order value, but it is only a single, moving point in the monetary value landscape. The gain or loss in each category of monetary value is the revealing information. The relationship between that categorical information and advertising, product, service, time, and price is the knowledge and wisdom to be gained from monetary value analysis. Reconciliation, then, must provide for total customer number counts in all dollar level monetary value categories. The numbers of customers added to or moved between all categories must reconcile with the total number of customers as defined by the reconciliation of recency counts. When the total number of customers by recency equals the total number of customers by monetary value, the analysis is in synch. Knowing the counts in each monetary value category and in total for today is completely useless unless you can replicate the reconciliation tomorrow and be able to examine and draw conclusions about the changes. The ebb and flow of customer movement between categories of monetary value is meaningful only when studied over time. For that reason, the discipline of reconciling and tracking monetary value on a regular, described schedule is necessary. Depending on the business, reconciliations can be accomplished daily, weekly, monthly, quarterly, or annually. For most direct marketing companies, a quarterly reconciliation is the minimum that can be expected to produce meaningful information; monthly is better. Over 20 rolling quarters, or 60 months, the display and charting of monetary value flow and the trend in direction develops a unique perspective of the direct marketing — indeed, any marketing — business. Monetary Value and List Rentals Monetary value as a component of the overall customer list is a valuable list enhancement. Companies renting your customer list will frequently specify dollar levels of monetary value when placing their list rental order. As an example, a list renter may come to you and ask for rental of all customers whose last purchase was $200 or more. If you classify monetary value in $25 increments beginning at $150 and extending through $300 and over, you will collapse the four segments ($201-$225, $226-$250, $251-$275, $276-$300+) into one monetary value segment ($200+) and aggregate all of the customer names onto a magnetic tape for the renter’s use. It is likely that your own list rentals from other companies contain specifications as to logical monetary value levels or ranges. If your average order value is $250, it is logical to want to rent names of buyers who have spent $200 or more; it is not necessarily logical to rent names of people who have spent only $8 on their last purchase. To demonstrate the value that exists with properly engineered monetary value analysis, consider the following request: "Give me all customers who have spent $200 or more on only one line item." By segmenting the customers in this manner, it is possible that customers with a higher dollar level purchasing authority can be isolated. That may be an important attribute for stimulating list rental income. The problem, of course, is that an adequate number of customer names must exist at the $200 dollar, one item specification. If only five such customers exist, there is little utility for either the renter or the rentee. Most likely, any customers having spent $200 or more will subsequently be rented. But for internal use, micro-segmentations of monetary value can often be quite productive. Knowing exactly how many customers exist in micro-segments of monetary value is one of the justifications for expanded dollar level segments and for frequent reconciliations. Whether you rent customer names with monetary value segmentation is not the important thing. The fact that monetary value segmentation will, in all probability, be an increasingly requested list enhancement is reality, as is the potential benefit from internal micro-segmentation of monetary value. Being able to segment and classify customer purchases by monetary value requires that line-item data be manipulated. If the database system does not allow for dollar value manipulation, line item count, and line-item product identification, only a small portion of the benefit of monetary value analysis is possible. It is not enough to know how much was spent. Knowing what was bought, how many were bought, the value of each item, and the value of the whole transaction is essential. Monetary Value and Advertising Frequency How often you mail, call on or telephone customers is, in part, determined by how much money those customers spend with you. If they spend a lot, you contact them a lot; if they spend only a little, you contact them less. Contacting customers frequently who spend too little is a problem; contacting customers infrequently who spend a lot is also a problem. Getting the mix right is the idea. Previously, it was concluded that the most recent customers deserve more frequent contact. Also it was concluded that the most frequent buyers deserve an increased number of contacts. And it was demonstrated that only by knowing the monetary value can those conclusions be valid. The customer who spends the most money should be contacted the most. If that customer is also the most frequent buyer and the most recent buyer, then you have isolated the best customer you have. If you can increase every customer’s recency, frequency and monetary value, you are turning every customer into best customers. If that is true, you can stop reading right here because you have the ultimate money machine. If that isn’t true, you may want to continue on to see what you have to do to get there. It is possible that the customer who has spent the most money is not a recent customer or a frequent customer. In fact, it is possible that the customer who heads the monetary value hierarchy bought only once five years ago. Should you send that customer a catalog a week for five years? The answer to that question depends on the size of that customer’s average order value. If the monetary value of the one order is big enough, it may be worth spending the advertising money to assure that you get the next order in five years. If the cost of the advertising is more than the net margin from the order, then there is little sense in sending five year’s worth of catalogs. The decision is, again, formulaic. Describing a customer in RFM terms can help. For instance, consider the following description of the top one year customer: One-year recency; 12-time frequency; $300 average order; 60 total aggregate orders; $18,000 total aggregate purchases. Numerically, this customer is described as: 1 – 12 – 300 – 60 – 18,000 Narratively, this customer is described as: A one-year recent customer, who buys monthly and spends an average of $300 per purchase, who has 60 previous orders and has spent a total of $18,000 over the past five years. If you simply add the numeric description together, the product is: 1 + 12 + 300 + 60 + 18,000 = 18,373 Divide the product by 1,000 and you arrive at a descriptive customer RFM score of 18.373. Array the RFM scores of all one year customers in a descending hierarchical listing and you have a simple but useful method for determining the relative comparative values of all one-year individual customers. Total the individual RFM scores for all one-year customers and divide by the total number of customers and you have the average RFM score for the one-year customer base. Compare that score month to month and you will be able to determine which direction the business is going. Based on the range of RFM scores, you can segment the hierarchical rankings and assign advertising mailing frequency accordingly. For example, all customers with RFM scores between 17.400 and 18.373 and above will be contacted 52 times a year, either by mail, telephone, in person or some combination or variation of those marketing efforts; customer groups scoring between 16.400 and 17.399 will be contacted 36 times a year, and so forth. The actual number of advertising contacts is only relevant to the business doing the advertising. Whether the total number is 52 or 4 is driven by the individual business and makes little difference in this discussion. What is important, however, is the recognition that some experiential hierarchy of advertising allocation is essential to success and RFM is one way to determine that strategic hierarchy. Without it, you are forced to guess at how many times to mail to a customer. Monetary Value and Customer Attrition The one customer that you don’t want to lose is the one who is spending the most money. Obviously, knowing who that customer is gives you an advantage in developing customer retention strategies targeted to keeping that customer happy, loyal, recent, frequent and profitable. Any one-year recency customer with high frequency and high monetary value who slips into the two-year recency box should be identified for immediate customer retention efforts. Even better, any one-year recency customer with a high RFM score who drops in overall scoring should be identified for immediate customer retention efforts before slipping into the two-year recency box. Remember, you never know when you have received the last order from a customer. Highly advanced customer retention strategies have one level of effort for slippage on recency alone, a second level of effort intensity for slippage of recency with frequency, and a third level of effort intensity for slippage of recency, frequency and monetary value. In other words, you work hard to keep one-year recency customers; really hard to keep one-year recency customers with multiple purchases; frantically to keep one-year, multiple buyers who spend big dollars. Unless the monetary value component is known, the customer retention effort is limited to recency and frequency; as has been shown, that is not enough. The high-dollar customers can fall through the cracks. One way to guard against high-value customer attrition is to isolate high- scoring RFM customers, as above, and pinpoint them as a part of the RFM reconciliation process. When high RFM scorers move from the one-year recency box to the two-year box, flag them for retention investigation and follow-up. Monetary Value and Product Propensity What products create high monetary value? It is more than likely that 20% of the products are responsible for 80% of the high monetary value. The 80-20 rule is almost always alive and well. In fact, somewhere out there is one customer who will buy enough of one product to equal your total sales from all customers and all products; it’s just a matter of finding the customer and the product. So far, no one has. Product commonalities exist in the higher regions of RFM scoring. You may also find that product commonalities exist in the nether regions of RFM scoring as well. Good products make good customers; bad products make bad customers. Understanding which products are which is the first step to identifying accurate product trends within the customer base. High monetary value is good provided that enough products are sold at that value to make it worth offering the products to customers. It does no good whatsoever to sell four items a year with high monetary value if the costs to stock, maintain and sell those products exceed the margin generated regardless of how much the sales are worth. The mark of an inexperienced marketer is falling in love with a product that doesn’t sell. The hierarchy of monetary value compared with product is a most valuable comparison and one that is worth the time and effort to isolate. Knowing which products drive monetary value is so basic as to be elementary; yet most direct marketing companies do not have a solid understanding of this relationship. A dispassionate divorcing of the number of orders, the number of units and the monetary value of those orders and units must be accomplished if this analysis is to have any utility for strategic purposes. Separate the orders, units and value in the mind in order to achieve objectivity. It is not enough to know only which products produce high monetary value. Knowing the units and the number of orders relative to that value is the essential knowledge to be learned. Together, those individual knowledge components produce a wise understanding of the marketing dynamic. A secondary understanding of product propensity is gained when high monetary value product sales are examined alongside associated product purchases. As an example, only one of Product A is sold but it produces multiple 1,000 unit sales of associated Product B in every case. An example is VCR equipment. It is likely that only one VCR unit is sold, but that unit produces multiple sales of video tapes. Once associated sales are added to the high monetary value sale, the marketing strategy can be radically different. Among catalogers and direct marketers with more than a single product, price is a primary element of monetary value. The higher the price; the higher the monetary value. The concept of price thresholds enters into the discussion of monetary value. One of the most frequently asked questions in direct marketing concerns the level of mark-up applied to price. While the percentage mark-up varies from product to product and industry to industry, and can be as high as 700% or more, rarely is profitability sustained at a mark-up under 250% in direct marketing. A two-time mark-up is too thin to carry the advertising costs. There are exceptions, but, as a general comment, direct marketing profitability begins at the 250% mark-up threshold; even there, profitability is elusive. If you have ten products at a 200% mark-up and ten products at a 400% mark-up, and you sell equal quantities of each product, you will end up with an overall 300% mark-up. Knowing the threshold for mark-up required to carry your business is the first step in determining price and, subsequently, monetary value. The second threshold is dollar value per product. Successful direct marketers establish a minimum product selling price that a single product must meet and they never vary from that requirement. After calculating the cost — to the penny — of processing an order, a direct marketer might come to the conclusion that no 2product can be offered that sells for less than $25. At a 250% mark-up, the cost of the product is $10; the cost to process the order is $8; this leaves a $2 contribution to overhead. To produce a $1,000,000 net contribution to overhead, 500,000 orders must be obtained. If, however, the minimum product price is $50, the product cost is $20 at a 250% mark-up and the cost to process the order is still $8, the net contribution to overhead per order is $12, and only 83,333 orders must be obtained to drive the $1,000,000 net contribution to overhead. At a 500% mark-up, only 41,666 orders are required to deliver the $1,000,000 net contribution. The higher the minimum product price threshold and the higher the minimum mark-up threshold, the higher the net contribution given the same number of orders. The converse is the higher the price threshold and the mark-up, the lower the number of orders required to deliver the same contribution to overhead. To achieve the desired monetary value and the desired mark-up, some mix of product is usually normal. High mark-up, high-priced products offset low mark-up, low-priced products to produce an acceptable average mark-up and average price. It is when the skew is toward the lower-priced, lower mark-up products that difficulties begin. Either the skew must be corrected or more orders must be obtained. No other choices exist. To understand all of this, every product must have its own individual history of number of orders, number of units, individual order and aggregate dollar sales, and mark-up performance. That information must be examined for shifts in trend on a regular and replicable periodic basis. To make it even more complicated, that same information must be examined relative to recency and frequency and determinations made as to the trends in one-time buyers, multiple buyers, one-year repeat customers, three-year repeat customers, and a variety of other RFM combinations viewed in light of product propensity and price. Another aspect of monetary value and pricing is the ability to study effects of price changes. Knowing what happens to overall monetary value when a 5% across-the-board price increase is levied is an important piece of information; equally important is what happens to monetary value when a 5% price decrease is made. Direct marketing is to some extent illogical. At times, sales and orders increase when prices are raised. Test a product at $99 and $129 and, strange as it may seem, the $129 price delivers better results. Perceived value is at the center of this illogical conundrum. Knowing how many are selling and how much revenue is coming in is the first step in comparing the effects of price changes. With a 5% across-the-board price increase, monetary value can be expected to rise by 5% as well. If it increases by only 2%, then 3% of the dollars are likely being withheld as a result of price resistance. That 3% will have to be made up in either additional orders, product cost savings or additional operating efficiencies. The ability to compare price effects on an aggregate and an individual product basis is essential in managing the mix and the thresholds for maximum profitability. Monetary Value and Advertising In a perfect world, it would make sense that some logical relationship between advertising and monetary value should exist. The products that produce the largest amount of revenues and the largest amount of margin should receive the largest amount of advertising expenditures of space and money. In fact, some allocation of advertising space and money should be made based on a descending hierarchy of monetary value. The product delivering the greatest value gets the greatest allocation; the product delivering the least value gets the least allocation. Of course, it is not a perfect world and almost never happens that way. Returning to our Star, Cow and Dog classifications mentioned earlier, there is sound reasoning for assigning advertising costs based on performance. A product or a product combined with recency, frequency and monetary value should, in all probability, be advertised in a manner whereby space and revenues are maximized. Square-inch analysis, albeit a topic destined for another book, is a formulaic allocation of total square inches of advertising space based on product monetary performance. Another way of allocating advertising is revenues received per dollar spent. Regardless of the method used, the relationship between products, advertising space cost allocation and profitability must be focused on and repeated — using oranges and oranges — regularly over time. And to accomplish this, you must have complete and accurate monetary value information on each product over time. The number of catalogs or other direct mail contacts to be allocated per customer is driven, in part, by monetary value. Remember our hardware customer example? Perhaps in the final analysis the customer who spends the greatest amount of money is the customer who gets the greatest amount of advertising and sales effort. But basing strategic decisions on monetary value alone is dangerous. It can produce enormous waste and it can lead to loss of opportunities. Monetary Value and Why It Isn’t Enough Having only the knowledge about the dollars spent is like knowing the balance in your checkbook but having no idea to whom or when you wrote checks. Overall, the balance may be accurate, but you may have paid some bills two or three times. There is no way to check what you are doing. And that is precisely why monetary value alone is not enough to successfully manage direct marketing performance. If you had a listing of all customers hierarchically ranked by total dollars spent, it would mean little unless it was augmented by a cross-listing of both recency and frequency. You have no idea whether those customers are still active (recent) or whether they are multiple buyers. As the three components — recency, frequency and monetary value — have evolved in the first three chapters, it should be clear that recency must be optimized with frequency and that resulting optimized hierarchy must then be further optimized by monetary value. The customers that rise to the top of the three-way hierarchy are the very best customers that you have. What you have done is shuffled the cards in the shoebox and isolated the best customers to be called on during the next trip. Optimizing Recency, Frequency and Monetary Value In the preceding chapters we have examined recency, frequency and monetary value in some detail and demonstrated the necessity for combining all three into an optimized hierarchy that isolates the most recent customers who buy most often and spend the most money all the way through the hierarchy to the least recent, frequent or valuable customers. Armed with that knowledge as a baseline, you can begin to overlay other optimizing characteristics on those areas of the customer list that hold promise. In this chapter, we will always assume that the three-way, optimized, hierarchical list has been created. At the top will be the very best customers and at the bottom will be the very worst customers. If your company can create this simple optimization of recency, frequency and monetary value, it is likely that it is among no more than 10% of the direct marketing companies in North America. If you can add any of the additional optimizations described in this chapter, it is likely that your company will be among only the top 1% or so of all direct marketing companies. Business to business marketers are familiar with Standard Industrial Classification (SIC) Codes. The federal government assigns every industry and sub-industry a unique code number so that all businesses can be classified in an orderly manner. By matching top performing customers against their industry SIC codes, direct marketers can see which industries offer better prospecting potential and can purchase compiled lists based on those SIC codes. It is possible to append the industry SIC code to a business to business customer list by passing the list across the Dun & Bradstreet master file of businesses. It is likely that about 70% of the customers will be "tagged" with SIC codes in this process. Periodic passing of the customer file will classify new customers and catch recently updated information on the 30% that are unclassified. Over time, valuable SIC information can be developed from the customer base. Of all the enhancements for business to business customer information, the SIC code is among the most important. Both business to business and consumer direct marketers can benefit from overlaying geographic elements on the results of RFM analyses. Knowing where the best customers come from can often be enlightening and can be used in the targeting of new customer prospecting. Geographic overlays can be done on a broad basis, such as regions or individual states or on a specific basis, such as ZIP codes or carrier routes. Isolating the best customers in a hierarchy by region, state or province, ZIP or postal code and carrier route may provide patterns and insights as to optimal customer location. Subsequent list rentals can then be structured to target these optimal geographic locations. Sophisticated software programs exist that cluster optimal response and purchase performance down to individual blocks and display these "hot cells" on detailed urban and suburban computer-generated maps. By picking off "hot cell" customer addresses and bracketing those customers, mailings can be developed that penetrate to the block level and reach neighbors one, two, three or more addresses surrounding top customers in all directions. With the advent of electronic interconnectivity and interactive marketing, it may be possible to create geographic "maps" of purchasing based on the interior of homes. If certain types of purchases are made by adults from the family room, only marketing messages targeted to adults will be directed to the family room through the fiber optic pathway. Purchases made by children from bedrooms will result in marketing to children via the bedroom fiber optic portals. Multiple marketing messages may be delivered simultaneously to different portals within the same house, all based on "micro-mapping" of response. Extending optimization and geographic mapping to the logical construct of "mind geography" and "mind-mapping," we envision the interconnectivity between the marketer and the thoughts of the consumer. With a "Friends and Family" interconnectivity, it may be possible to do all of your Christmas shopping and assure that no duplicate presents are purchased, or that every article of clothing is the right size based on knowledge contained on the sizes of all family and friends in the participatory electronic mapping database. For example, one day the hybrid TV- computer-fiber optic portal may tell you, "Buy a size 14 for your mom instead of the size 12 you ordered. She has made two purchases in the last week in size 14 for herself." Or, "Your sister bought a blue tie for your father for Christmas. His profile indicates he is interested in an ivory, cotton sweater in size XL." Once the mind of the customer, the desires of the customer and the inter-relational thoughts and desires of friends and family are cyber-mapped, it will be possible to coordinate them in an elegant dance of cyber-marketing far more efficient–indeed, almost prescient–than today. The optimized, geographic cyber-mapping of the thoughts of the customer relative to purchase potential is a logical outgrowth of technological advancement. Optimized RFM Plus Financial Information In the same way that geographic location can be mapped and top customers isolated relative to geography, so can financial information be optimized as an overlay onto RFM. Annual income, annual sales or other financial indicators can be structured hierarchically against the ranked customers and examined for patterns and commonalities. Once optimal financial levels of the best customers become apparent, tailoring of prospect list rentals can be enhanced to target those prospects who are financially most like the top performing existing customers. Optimized RFM Plus Gender Knowing the gender of the best customers is essential for an optimized marketing strategy. Whether business to business or consumer, the creative approach may logically differ; so may the product propensities. If the top performing customers are female and the creative or product is designed for males, or the other way around, there may be a lack of profitable efficacy in the creative execution and product conception.
Once the business to business customer base has been optimized by recency, frequency and monetary value, it can be helpful to then rank the companies by numbers of employees. Correlations may exist between top performing customers and a particular size of company. Generally, employee size is captured in categories such as 1-5, 6-10, 11-25, 26-50, 51-100, 101-250, 251-500 and 500 and larger. Other categories are also used; it makes little difference what level of detail you capture as long as the detail is meaningful. For example, if only employee size above 500 and below 500 were captured, the information would have little value in selecting segments from business to business mailing lists. Knowledge must be digestible. Knowing that the greatest portion of the best customers come from businesses with 50 or fewer employees is much more useful and reliable information. Armed with that understanding, appropriate cross sections of mailing lists can be selected that zero in on businesses with 50 or fewer employees. As can be expected, it is likely that a bell curve will form when customer performance and employee size is examined. At the extremes will be the very best and the very worst customers with the bulk of the average customers at the bulge. Knowing how and why an average customer becomes an extraordinary customer is the essential knowledge to be gained. Optimized RFM Plus List Source From what mailing lists do the bulk of the very best customers come from? By overlaying and segmenting the optimized hierarchy with the mailing lists, a picture should emerge that reveals common list sources. Relying purely on response statistics is not adequate for evaluating list performance. High response lists may produce terrible customers; conversely, lists with horrible response may, perversely, produce the best performing customers. Optimized RFM plus list source coupled with the exact cost to acquire a customer and lifetime net profit of that customer are the cardinal pieces of information that must be obtained. Optimized RFM Plus Title Business to business customers have titles. By capturing titles of customers, some relationship can likely be identified between title and customer performance. If a large percentage of the top performing customers have the title Warehouse Manager, it is a safe bet that you should be looking for more warehouse managers as prospects; you may also benefit by mailing to the warehouse managers of existing customers. You might add a second customer inside a company that is already buying from you. Title is often a function of company size. Small companies with few employees generally do not have Vice Presidents of Purchasing or Human Resource Directors. In fact, these companies generally have Owners and Others. After optimizing RFM by employee size, optimizing by title can often prove beneficial. Coming as close as possible to the person who is likely to be the best customer inside a company is an important objective when designing advertising campaigns and mailing schematics. If your analyses determine that warehouse managers are, indeed, your best customers, then by all means add the title "Warehouse Manager" to a portion of your prospect lists and test the results. If that group is significantly more profitable than nontitle prospect mailings, you have discovered a meaningful piece of knowledge. In some areas of direct marketing, such as insurance, age is an integral part of the product. Elsewhere, age can be a by-product wherever date of birth information is captured. An optimization of customer performance with age may reveal productive information. If the very best customers tend to be between 45 and 55 years of age, that is strategically vital information to know as it allows you to focus on prospects who fall into that specific age bracket. Knowing the influence of age on profitability is also essential for determining which nonprofitable customers are likely too old or too young for intense marketing efforts. A mediocre customer who is about to enter the "Golden Bracket" may be a better bet than a mediocre customer who is 20 years past the prime customer age. Intricate relationships of potential begin to make themselves evident when actual customer performance and potential customer performance are analyzed alongside customer age. Business to business direct marketers may find correlations between customer performance and the number of years in business. Further optimizations of top performing customers with years in business, employee size and annual sales can produce a profile or indication of stability. As you can see, numerous possibilities exist for enhancing the optimized recency, frequency and monetary value information. By adding optimization upon optimization, a smaller and smaller slice of the customer base is isolated. For some companies, multiple optimizations are unnecessary and would add little benefit. For other companies, the ever-sharper definition of the customer is essential to survival. Much depends upon size. A fully furnished RFM optimization program is beneficial only if sufficient customers exist in the database to assure statistically valid results. An optimized hierarchy of the best customers that contains only two top performing customers is of little or no practical value. But where sufficient numbers of active customers exist to produce meaningful and actionable slices of performance excellence, optimal RFM efforts are useful. Of all the information available to direct marketers in the ever-increasing aggregation of information, recency, frequency and monetary value remain as the cornerstones of a well-constructed direct marketing house. Hopefully, some, part or all of the information in this book will assist you in managing your direct marketing business and will add to a strengthening of our proven free-market, entrepreneurial spirit of commerce. Copyright 1992-1998 by Donald R. Libey. All rights reserved. No reproduction or use of this material may be made by any means without written permission from the author, Donald R. Libey, Libey Incorporated, 1308 Keswick Avenue, Haddon Heights, New Jersey, 08035, U.S.A.; telephone: 609.573.9448; facsimile: 609.573.9685; e-mail: libey@bellatlantic.net. Copyright abuse will be aggressively pursued and can result in judgments awarded by the courts of up to $100,000 or more to the copyright holder. Advertising 1, 60, 73–76,
82, 84, 89,
90, 96, Attrition 20, 21, 42–46, 116, 118 Characteristics 5, 53, 58, 98, 127 Database 3, 10, 14–18,
21, 24, 28,
29, 42, Frequency 1, 1–5, 28,
50, 51, 53–95, 97–101, Lapsing 91 Marketing 1–6, 10, 11,
14, 16–20,
25, 28, Monetary 1, 1–5, 28,
50, 51, 53,
70, 82, 95, Pricing 61, 63, 72, 78, 106, 120, 122 Product 10, 12–14, 32,
54, 56–61,
64–67, 71, Quantity 38, 39, 59, 61–63, 73, 74,
77, 92, Recency 1, 1–14, 16, 18,
20, 21, 23–30, 32– Reconciliation 30–34, 42, 52,
82–84, 109– Recovery 46 Rentals 19, 34–36, 88, 111, 112, 129, 131 RFM 5, 53,
98, 115–118,
122, 127–129,
131– Segmentation 14,
18, 21, 24,
30, 31, 33–36, Time 3, 4,
6, 8, 11–14, 16, 17,
19–21, 23–25, |