Retention in a High Churn Biz Model
# 63: 1/2006
Drilling Down - Turning Customer
Data into Profits with a Spreadsheet
Customer Valuation, Retention,
Get the Drilling Down Book!
In This Issue:
# Topics Overview
# Best Customer Retention Articles
# Retention in a High Churn Biz Model
Hi again folks, Jim Novo here.
How do you handle customer retention in a business model with
built-in uncontrolled churn? By improving customer acquisition
practices, that's how, and the topic of this newsletter.
We've also got a couple of great customer marketing links, including
one on Six Sigma Marketing
In case you were wondering, there's wasn't a December newsletter,
as I was up to my ears in producing the collegiate-level web analytics
courses for the Web
Analytics Association. If you want to learn more about the
courses (or take them), more info can be found here.
Let's do the Drillin'!
Best Customer Marketing Articles
January 19, 2006 CMO Magazine
This is a fantastic piece on a large scale Six
Sigma Marketing implementation at Xerox. The bottom line: it
can be done and you could save millions doing it.
*** The Secret Sauce
December 27, 2005 Internet Retailer
This is a pretty well organized article on the different facets of
conversion. Nothing earthshaking, but if you're having problems
putting everything you have heard about optimizing conversion in
context, it's for you.
Questions from Fellow Drillers
Retention in a High Churn Business Model
Q: I bought your book a few months ago and I thought it was great.
A: Thanks for your support and kind words!
Q: I did what I could using your book, but my mind is going in circles and I need help. Unfortunately I can't pay big bucks and am hoping that the "case" will be of enough interest to allow you to help me out at less than you could charge to bigger companies.
A: Perhaps, I am always open to interesting projects if I can, for example, use them to publish a case study (names changed to protect the innocent)...
most large companies do not want customer retention data revealed and
it is tough to get case studies out there.
Q: What I do: My company sells prepaid calling cards on the web.
I have been doing this for 2 years and 4 months now. Consumers purchase the service to place international calls at reduced rates from their landlines and cell phones.
They have no obligation to re-purchase. They can buy any increment between $10 to $100 at a time. They can recharge their accounts at any time.
Q: What I've found so far:
They seem fall into two main groups those that buy only once, and those that buy more than once, but then each of these sub-groups has lots of sub-segments and it
is confusing. Some last a week, others last 2.4 years.
A: This is typical...
Q: My goals:
Improve customer loyalty by flagging deviation from aggregate customer "norm" in their group as well as the customer's own "norm" based on that customer's previous behavior. Campaign automation via trigger-points.
Determine a "safe" customer acquisition cost by targeted source.
A: Sounds like you read the book!
Q: What the problems are:
* I need someone to help me hash out the RFM parameters to use in my calculations,
i.e. "what bean am I counting."
* I've had some trouble with the Frequency concept. If a guy buys $100 and only does it once a month, is he worse than a guy who buys $10 eight times a month?
Both Recency and Frequency are skewed here towards the small increment buyer.
I could use actual "last calling card use date" as the parameter but there seems to be problems here as well.
I need help with coming up with the right set of parameters or proxies to track.
* some Present value and future value issues related to current customer age.
I can go back over 2 years here so the data is available. Based on lots of excel work, I think I have a handle on this but want to run it by someone who knows
more than I do.
Well, first off, I'm not sure in this biz that RFM is the right customer behavior model to use.
The Latency model is probably the more appropriate one, as in "change in rate of usage". You have (I think) superior behavioral information - better than most businesses - so there is probably a better way to look at this then using the RFM model.
Rate of change is a very powerful predictor of future behavior in a nearly frictionless business like this one.
For example, let's say usage from sign-up increases 10% the second month, then 20% from there the 3rd month, then 30% from there the 4th month. Then in the 5th month, usage increases only 20%.
Even though this is an increase and would normally be viewed positively, the *rate of increase* has
slowed, and this is a predictive warning flag.
On the "front end" (acquisition), I would look at which campaigns new customers come from and how long those customers last / their spend by campaign source.
In this kind of business where there is a high uncontrollable churn rate anyway, this is your most effective
Reallocate ad budgets away from campaigns generating low value customers towards campaigns generating high value customers. Please be aware of a direct marketing truism - often the higher the response rate to a campaign, the lower the long-term quality of the customer generated.
When you are in a "commodity" business like this one, retention is a very tough game, period.
So if you act upfront by tuning acquisition, you will be ahead of the game already.
Smart acquisition = better retention.
On the back end (retention), I would try to find usage patterns that predict defection.
In other words, aggregate high value customers that have defected and look for patterns that predict this defection.
Typically, this is done by start date, e.g. "all the new customers from Jan 2004 that spent over $1000 in 2004 then cancelled the service".
What is it about their behavior that is similar? Did they come
from the same source? How do their usage rates change,
especially before the defection?
For example, what was their usage rate going into the cancel?
Did the usage ramp up continually month after month post sign-up, only to reach a peak and start to fall off leading into the defection?
This is a typical pattern and one that could signal the need for marketing at the point where usage deceleration is
detected - this deceleration in usage becomes your "trip
wire". Of course, you need to make an offer of some kind to keep them, and (I'm guessing) the per minute rate is probably not something you have a lot of room on, so extra services that really cost very little to provide (caller ID?) that you may charge for
now could be offered free to encourage the customer to
I checked out your web site and you obviously have a keen understanding of the market and your positioning in it.
The whole FAQ approach you take to explaining phone rates and the deception that occurs in this business (hidden charges, etc.) is fantastic.
For example, explaining all the scenarios where your rate might not be lowest is a great
seller piece of confident seller communication.
So I think you are doing a lot of things right, but you are in a business that suffers from a high churn rate as a "given".
First concentrate on where you are allocating marketing resources to optimize the front end - acquire customers that are more likely to stick in the first place.
Then look into retention and what you might do to target best customers who exhibit behavior that likely leads to defection.
For larger companies in a commodity business, the best answer to retention issues is usually a *compelling* loyalty program.
Many of these programs are poorly executed and provide little relevant value, so they don't really work.
An example of an approach that does work can be found here.
You are obviously a smart cookie on the marketing side, so perhaps you can find some inspiration in this case as far as offering targeted benefits to retain best customers. Good Luck with it, and if you have any more questions on topics from the book, let me know.
You're a customer now!
If you are a consultant, agency, or software developer with clients
needing action-oriented customer intelligence or High ROI Customer
Marketing program designs, click
That's it for this month's edition of the Drilling Down newsletter.
If you like the newsletter, please forward it to a friend! Subscription instructions are top and bottom of this page.
Any comments on the newsletter (it's too long, too short, topic
suggestions, etc.) please send them right along to me, along with any
other questions on customer Valuation, Retention, Loyalty, and
'Til next time, keep Drilling Down!
- Jim Novo
Copyright 2006, The Drilling Down Project by Jim Novo. All
rights reserved. You are free to use material from this
newsletter in whole or in part as long as you include complete
credits, including live web site link and e-mail link. Please
tell me where the material will appear.