New RFM: Predicting Student Churn

Jim answers questions from fellow Drillers

Topic Overview

Hi again folks, Jim Novo here.

Only attempt to control what you can actually get control of. In a massive operational system, sometimes everybody agrees they would like to improve this or that customer retention metric. But the reality is, because of the way the system is structured / managed / operated, it will be nearly impossible to improve the specific metric. Such is the case with a University. So you gotta carefully evaluate and choose what metrics you think you have a chance to control. Sound difficult? Sure it is. But difficulty hasn’t stopped us yet, has it fellow Drillers?


Q: It is clear that retention of students is a complex issue. The students’ satisfaction with the university will be partly determined by their experience during their first semester with the university. I have identified that each service encounter will contribute to the overall impression that the student has of the university. Some encounters are ‘moments of truth’ and will have a major impact on the student’s perceptions of the university.

A: Hmm, interesting.

You just got very lucky, I happen to have first-hand experience on this topic, which is very rare, as not many educational institutions are thinking this way. You should be congratulated for making this connection, though it will be a difficult battle dealing with the university administration on making changes, in my experience.

Q: I would much appreciate if you could advice me on the retention strategy and what approach the university should take to retention. Also, any ideas on management of moments of truth, particularly what enhances and detracts from the customers’ encounters with the university.

A: Please consider this old business maxim: Only attempt to control what you can actually control; otherwise you will end up not having an affect on anything.

It very well may be that the various “touchpoints” exist and can be defined, but can you reasonably control any of them? Which ones, and how will you control them? This is where you need to focus your efforts.

In my experience, a university is not the kind of place where you can undertake a “customer service education program” with employees and expect compliance at all the touchpoints. So what you have to do is pick the major points of influence where you know you can exert some control and seek to prove your case with facts and testing.

Continue reading New RFM: Predicting Student Churn

Customer Segmentation: Tangible vs Intangible Cost, Let Data Define Segments

Jim answers questions from fellow Drillers

Topic Overview

Hi again folks, Jim Novo here.

This month in the newsletter we answer questions on the nitty gritty of the actual discovery work by taking a very deep look into the whys and hows of segmenting customers. Straight-up and to the point, put on those data shoes and Let’s do some Drillin’!


Q:  Hi Jim, I’m a great fan of your work!

A:  Well, thanks for your kind words.

Q:  I have a basic question for you.  We are an online retailer and thus use email as the primary marketing communication channel (we do use Direct Mail to our best customers around holidays).

A:  Those are smart choices.  I’ve seen some stats on using direct mail to drive lapsed online customers either back online or into a store that are very encouraging, real money-makers for retail.  Definitely worth testing, though in both cases, the product mix averaged higher ticket than your category typically does.

Q:  However, we don’t have a set customer segmentation technique and thus no specific customer segments.  One outside consultant, a statistician, had suggested looking at a new customer’s activity in the first 30 days and then classifying them into High Spender, Frequent Transactor, etc. segments.  Not sure how well it works.

A:  That’s quite unusual, I think.  It would work in the first 30 days, but I think you would have to re-classify every 30 days using a scheme like that.  Considering web-only behavior, the typical retail lifecycle beyond 2nd purchase (many buy only one time) is a ramping to a peak and then a more gradual, but still steep, falloff in purchases.  The model above would not take this into account, and while the initial label might be accurate, it soon would not be.  That’s not to say these kinds of models don’t work, but it usually takes years of testing and study to perfect them.  “Data miners” often believe the numbers will simply tell them things like this, but they don’t take into account the human behavioral and other mitigating factors which may not be in the data.  

For example, Recency and Latency are really “meta-data” about customer behavior; they are data created from other data.  You can’t just look at the first 30 days of transactions and give a customer a label; customers have LifeCycles and you drive the highest ROI when you take advantage of knowing these cycles and acting on them to increase profits.

Q:  I feel that we target our customers primarily by their category purchases, and not by any kind of behavioral model.

Continue reading Customer Segmentation: Tangible vs Intangible Cost, Let Data Define Segments

Customer Perks Marketing

Jim answers questions from fellow Drillers

Topic Overview

Hi again folks, Jim Novo here.

What’s the best way to handle customized marketing programs, particularly if you are using customer value as the key segmentation approach? Surprise and delight, perhaps slightly influenced by making more money. Sound good? Let’s Drill it … 


Q:  Jim, do you have an opinion on overt versus covert customer benefits?  What I meant by overt vs. covert… have you seen clients do programs where they TOLD their customers they are a valued (Gold, Platinum) customer and provided tangible benefits, vs. others who have just covertly treated these customers specially in some way (i.e. priority routing, better reps, thank you calls, etc.)

A:  Well, a program won’t be very effective if everything is completely covert.  I mean, it’s nice to get great service and that certainly contributes to customer retention, but recognition is much more powerful.  The customer needs to know they are being treated specially at some level to maximize program effectiveness. Why?

Something like call routing is a good example.  If a customer is getting priority call routing and they don’t know it, they may think the service is good.  If you tell them they are going to get it and then they get it, it’s an entitlement they earned.  More powerful, and more effective in keeping the customer.  Let’s say they are thinking of defecting.  If they don’t know they are getting priority routing, they could suspect the service might be as good at the competition.  If they know they are getting priority routing, the question becomes “Does the other guy do this to?  And if so, will he give it to me?”  See what I mean?  It’s much more powerful for the customer to know they are getting special treatment than not to know.

Continue reading Customer Perks Marketing