Clumpiness and Customer Lifetime Value

Knowledge at Wharton

Ask about Knowledge at Wharton

Episode Description

📂 Marketing 🏷️ Research

Read the full transcriptHide the full transcript
  1. well one of the most established practices in the field of marketing and customer valuation is to summarize a
  2. customer using what's called rfm segmentation which means I take everything I know about my customer and
  3. I compute just three simple numbers how recently did they buy when's meaning when's the last time they bought how
  4. frequently do they buy meaning the number of time periods in which they bought and when they buy how much money
  5. do they spend that's called r FM segmentation it's the basis of the way in which most companies decide who are
  6. the valuable customers and who are the non-valuable customers and my research is very simple it basically says that's
  7. not a complete characterization of customers you have to add one more letter to rfm and I call that c which
  8. means clumpiness which means some customers do Buy in a regular pattern and historically if you bought orange
  9. juice if you bought diapers you bought things in a regular pattern but clumpiness refers to the fact that
  10. people buy and burst and those burst periods indicate something very different about the customer and that
  11. those customers could be extremely valuable the key takeaways of my research is very simple let's imagine
  12. you want to build a simple what I call a simple mathematical model you want to predict who are going to be the valuable
  13. customers in the future and you have four things you can use to predict it as I mentioned recency frequency monetary
  14. value and let let's say the marketing spend towards the customer those are the classic ways in which companies build
  15. what are called scoring models I'm claiming you need to add one more number and that's see how clumpy the customer
  16. is this is no more difficult to compute than rfm you can do it in Excel it's very quick to compute you can compute it
  17. for literally a 100 million customers in a second and the findings of my research
  18. suggest that higher clumpy customers are worth more out of sample meaning in their future value even after
  19. controlling for rfm and marketing expender which means we have found another variable that firm should track
  20. about customers and use it to predict their worth in the future two things surprised me about my
  21. conclusions one is um I just figured that this rfm based segmentation which had been around for so long and is used
  22. by so many firms had been validated in the sense that there wasn't any anything else simple out there that could help
  23. explain customer value I mean you can do all kinds of fancy web scraping and all
  24. kinds of other variable construction but clumpiness is so simple so first I was surprised that that had been missed that
  25. in other words hot and cold periods are indicative of something about the customer I think the second part that
  26. surprised me is that at least the data sets I've analyzed it's true for let's call it digital and online consumption
  27. Goods but it's not true for regular consumer package Goods in other words historical models I can see why they fit
  28. fine because you buy toilet paper in a regular pattern you buy orange juice in a regular pattern but you don't consume
  29. Hulu in a regular pattern you don't bid on auctions at eBay at a regular pattern
  30. you don't buy books at Amazon on a regular based pattern so I think the two things that surprised me is number one
  31. that people had missed what seemed to me to be something fairly obvious and secondly that it applied
  32. historically if you look at historically purchased Goods clumpiness really isn't
  33. there but if you look in the new wave the new economy clumpiness is pervasive in every data set I've
  34. analyzed I think of all the research I've done over my career which is now it's hard for me to believe but it's
  35. been a 20-year career I think this is probably the most practical thing I've done um the work I do tends to be what I
  36. call fancy complex statistical modeling and this is n about statistical modeling
  37. this is about a number clumpiness that firms can actually compute today they don't need to collect any additional
  38. data it's the same data they're using to compute R F&M and customer lifetime value and they can figure out how much
  39. value it adds to predicting customer value your rank ordering of customers will change your decisions about which
  40. customers are valuable to reactivate imagine customers have churned well which ones are valuable to reactivate my
  41. claim is the clumpy ones even though they've churned those are the ones to reactivate cuz if you reactivate them
  42. they'll come back and be clumpy again and do a lot of stuff in the future so I think it has huge practical value and
  43. the beauty of it is if you go to my website I have an Excel sheet there that has worked out examples it actually has
  44. an Excel sheet that you can just download and you can start using clumpiness today a lot of people have today talked
  45. about big data and I'm actually I love big data but I'll tell you what I love even more than big data I love data
  46. compression and what I mean by data compression is you can collect thousands and thousands of variables on people now
  47. you can track where they are you can track what they bought what web pages they looked at but that's not science
  48. that's a that's data collection now the question is which of that information is
  49. actually useful for the business problem at hand and that's what I call data compression so the way I view clumpiness
  50. as an addition to traditional variables like rfm marketing activity and stuff like that I Vis I view it as a form of
  51. let's call it increased data compression I'm just telling you you need to cover you need to keep a little bit more data
  52. you can't compress things down to three numbers you got to compress it down to four so I view This research that I'm
  53. doing as kind of I view it in the data compression world I love the problem of taking big data and compressing it down
  54. to small data and that's how I view clumpiness the part that's unknown to me right now
  55. which is I've done a lot of work on clumpiness I know it exists across Industries I know it exists I know it
  56. can be predict of predictive value here's what I don't know what causes it so what I do know is I've related
  57. marketing activity to clumpiness so firms can try to make you clumpy by sending you an email by sending you a
  58. catalog by targeting you Etc that much I know but I haven't really studied yet what's the optimal way in which firms
  59. should targeting you target you knowing that clumpiness exists I haven't looked at like for example do you consume more
  60. clumpy content if it's a series like imagine watching Breaking Bad or Mad Men or something like that or imagine you're
  61. a firm and you're trying to sell a sweet of products like you know a facial care
  62. line and a you know moisturizer line and all this other stuff should you package
  63. it together and make it seem like people are progressing towards a goal so here's
  64. what I do know I know mathematically how to compute it I know it's trivial for firms to do I know it's predictive but
  65. the part that's left unknown to me is the psychology of why which is why I'm partnering right now with a lot of my
  66. more consumer psychology oriented colleagues we're going to start running a lot of Behavioral experiments in the
  67. lab to try to get to the under underlying psychological underpinnings of why people behave in a clumpy
  68. fashion I'm not sure I've seen much about clumpiness if except what you see is you see stories in the news all the
  69. time about people kind of binging on content and so I like the word clumpiness other people like the word
  70. binging um the reason I like clumpiness is that it refers to you know the opposite which is non- clumpy which is
  71. kind of equally spaced kind of arrivals or equally spaced purchases so what I would say is I don't think I've seen a
  72. story about clumpiness but anytime you see a story about people binging content people consuming things you know a
  73. student sat up for 18 hours watching this it applies and the concept is so pervasive and every time I talk to
  74. whether it's managers students academics about it everyone believe it exists the
  75. part as you mentioned in your earlier question that shocks people is that it's actually predictive of customer
  76. value I think it dispels the idea that in some sense um customers can just be categorized by a simple set of numbers
  77. um you need to go a little bit beyond that you need to go a little bit beyond what I would call Simple theories of how
  78. people behave and what clumpiness if you actually think about what clumpiness says is if you look at recency frequency
  79. monetary value which is kind of the historical basis of consumer Behavior it basically ignores what I call the inter
  80. Ral times it basically says I can take all the data like if a two-day P it was a two-day window and then a 4-day window
  81. then a a three-day window then a sixday window I can throw all of that away and all I need to know is when's the last
  82. time you came and how many times did you come but what this dispels is that the arrival pattern of people is
  83. uninformative it's very informative people that come and burst then go away and then come back and burst and then go
  84. away it's just I think those are just different types of people I think those people are fundamentally different I
  85. personally believe there are clumpy type people and non- clumpy type people however what we've also shown is it
  86. varies by product categories so we found for example that women tend to be more clumpy than men we found that younger
  87. people tend to be more clumpy in their consumption than older people so I think what's really going to dispel I think
  88. what the myth that we're going to dispel is that like not only are all people created equal but that there are simple
  89. ways to just categorize all people into a certain type I think what everyone's done is
  90. there's a whole class of mathematical models that have been popularized although they've been around for 50
  91. years but have been popularized over the last 10 years called hidden Markov models the idea is very simple let's
  92. imagine there are two states of the world you're in a hot state or a cold State and you rotate back and forth
  93. between a hot and a cold state that mathematical model is clumpiness you're hot you do a lot of stuff you're cold
  94. you don't hot cold hot cold what TP what separates this work is the work I'm doing isn't Ivory Tower mathematics it's
  95. a simple number that someone can compute so I fit hidden Mark of models to data what I wanted to do was to bring to the
  96. practitioner a way that they could compute a simple number it's a statistic it's not a statistics paper it's a paper
  97. about a number a statistic as we call it you just compute the number and then do
  98. what you want with it you could try to use it to predict customer value you could use it to see are men more than
  99. women you could use it to segment people that's what typifies and separates this
  100. work is that it's a simple metric based approach that practitioners can use it's
  101. not a fancy modeling based approach but they're both trying to cover the same problem what we've studied so far with
  102. reaching clumpy customers is we've studied whether email cataloges different types of marketing channels
  103. are more effective and what we found not surprisingly is email has more of a short-term effect as you would expect
  104. catalog has more of a longer term effect what we've yet to really understand is are there certain words in a email or
  105. catalog or you know video campaign that will engage or you know if you'd like cause people to be more clumpy are there
  106. certain topics that are more clumpy more some product categories that will necessarily be more clumpy all we've
  107. done so far is I think I've established the phenomenon on exists I know it exists across lots of Industries I know
  108. certain types of people tend to be more clumpy the part that I haven't done which is shocking because I'm a
  109. professor of marketing is talk about the marketing implications of it yet that's
  110. going to require bigger and newer data sets that allow me to link things about marketing campaigns to people's clumpy
  111. Behavior I know how to do it it's just I need richer and better data to do it I'm thinking about three different
  112. streams to follow up this clumpy research first of all I'd be thrilled to just analyze more data sets and prove
  113. how pervasive the clumpiness measure is so I've analyzed data sets from Amazon from CD Now from eBay from Hulu from
  114. YouTube and also from some traditional consumer package Goods companies look if PNG wants to contact me tomorrow I'm
  115. happy to apply clumpiness to their work if Google wants to contact me tomorrow if Goldman Sachs wants to contact me
  116. tomorrow if fizer wants to contact me tomorrow I'm happy to work with their data and understand clumpiness and how
  117. it can predict customer value so that's one area I just want to apply it to new data sets the second is I want to
  118. understand the psychological processes why are people behaving in a clumpy fashion the third and final pieces I
  119. want to relate marketing activity to clumpiness now that's going to require not just people's behaviors like what
  120. did they do what websites did they visit what did they purchase but information about the marketing campaigns themselves
  121. possibly even the copy of the marketing campaign which channels they were sent through and that's going to allow me to
  122. come up with optimization ways for firms to optimize their marketing campaigns to
  123. activate clumpiness so those are the three Avenues I'm going to be working on [Music]
  124. next

Brief Summary

Michael H. Market discusses customer clumpiness and its impact on predicting customer value in marketing.

Key points

  • Understanding Clumpiness in Customer Behavior. Clumpiness indicates that customers buy in bursts, revealing their potential future value.
  • The Importance of Clumpiness. Clumpiness can change how companies predict customer value and worth.

Related episodes