The Web Retailing Example - Drilling Down Newsletter #
28: December 2002
Drilling Down - Turning Customer
Data into Profits with a Spreadsheet
Customer Valuation, Retention,
Get the Drilling Down Book!
Now also available online through
Amazon and Barnes & Noble
The Marketer's Common Sense Guide to E-Metrics - 22 benchmarks to understand the major trends, key
opportunities, and hidden hazards your web logs uncover. I wrote
this manual with Bryan Eisenberg of Future Now, the visitor conversion
To find out more about this topic:
In This Issue:
# Topics Overview
# Best of the Best Customer Marketing Links
# Tracking the Customer LifeCycle: Recency
# Questions: ROI on Multi-Step Promotions
Hi again folks, Jim Novo here.
This month we've got the usual "best of" Customer
Marketing article links, we explore the concepts of "new
customer" and "customer" with the owner of
IMissAsia.com, and roll up the old testing sleeves to dig down into
the depths of control groups, random samples, and halo effects.
If you're not in the data-driven mood, hey, read it next week -
it will still be here. Take some time off, relax and enjoy the
season. But perhaps you have a bit of time on your hands, as
business activity grinds to a halt. If you are in this
Let's do some Drillin'!
Best Customer Retention Articles
This section flags "must read" articles moving into the paid archives
of trade magazines before the next newsletter is delivered.
If you don't read these articles by the date listed, you will have to pay
the magazine to read them from the online archives.
The URL's are too long for the newsletter, so these links take you to a page with more info on what is in the article and a direct link to the article.
Note to web
site visitors: These links may
have expired by the time you read
can get these "must read" links e-mailed to
every 2 weeks before they expire by subscribing to the newsletter.
the Web's No. 1 Activity - Searching
Expires December 26, 2002 DM News
For those who have been paying attention, search marketing found favor
a long time ago. What better position for a marketer to be in
than to snag people searching for a product?.
Bolsters Analysis With Inhouse CRM
Expires Jan 5, 2002 DM News
Say it isn't so! People who signed up for a sweepstakes actually
turned out to be more profitable than customers? Helps if you
sell big ticket, as this retailer found out when they finally took
control of their data.
Tracking the Customer LifeCycle:
Real World Examples
If you are new to our group and want to review the previous
LifeCycle metric - Latency - that discussion is
along with the Real World examples Hair Salon and B2B
Software. The previous piece on Recency is here.
Recency: The Web Retailing Example
IMissAsia.com offers an e-mail newsletter on every page of the
site. The owner tries to create a broadly appealing piece,
mixing some new content with links back to areas of the site
experiencing high activity - specific discussion boards, products,
news clips, etc.
The owner has always felt the visitor / customer should drive the
direction of the site; if certain topic areas were getting the most
traffic, then those must be the most interesting or attractive topics,
and likely the ones with appeal to the most people. This
"swim with the tide, not against it" approach had always
worked well in the past as a driver for newsletter content.
Within this content, the owner carefully mixed contextual sales
opportunities directly related to the content, along with one or two
more aggressive product pitches. This formula had worked well
and the newsletter drove a good chunk of sales.
But the owner of IMissAsia.com was getting worried about response
to the newsletter, which has been falling. Perhaps swimming with
the tide was not a good idea, and the content should explore "not
popular" issues and products? Perhaps the product pitches
were too frequent and aggressive? Perhaps this market was just
slowing down because of the economy? Or worse, perhaps the owner
had already "creamed" this market and the best days were
One thing the owner knew for sure - the percentage of total sales
from new customers was falling. Now, this could be a good thing,
the owner thought, because it means more sales are coming from repeat
buyers. But it could be a bad thing, if what it means is the
market is saturated and the best days are over. How to resolve
this question? And how is the newsletter affecting this issue,
if at all?
The owner thought a lot about new customers, repeat customers, and
the newsletter. What is a "new" customer, anyway?
Are they new only the day they make a first purchase? Are they
still new if they haven't made a purchase 30 days later? 60 days
later? 6 months later? Do they have to make a second
purchase to not be "new"? When do they stop being new?
For that matter, when is a customer not a customer any more?
If they purchased twice or more and have not purchased again for 6
months, are they still a customer? What about no purchase in 2
years? 5 years? When do customers cease to be customers?
What does the customer base of IMissAsia.com really look like?
The owner realized the only way to answer these questions was to
actually look at the customer data, and to make decisions on what
these ideas meant for this business. Customer types for
IMissAsia.com probably would not be defined the same way as a
customer types for Boeing, Wal-Mart, Oracle, or Ford. No, these
customer definitions needed to be based on the facts of this
particular business model.
The owner also realized something else - if there are no
definitions, there can be no measurement. And without
measurement, there is no way to understand the dynamics of what is
happening to the business, for example, why the response rate to the
newsletter is falling. All the owner knows is one thing, the
"what" - response is falling. The owner wants to
understand "why." And there is no way to get to
"why" without understanding the "who" first.
Response to the newsletter is falling because not as many customers
are responding. Who is not responding to the newsletter?
Is it new customers? Is it repeat customers? Is it
"best customers"? The owner realizes there is no
definition of best customers either. If these things were defined,
the owner might be able to measure and understand what is happening.
Then another realization - not just defined, but tracked over time.
It does no good to define customers and count how many there is of
each type; what the owner needs to know is how these counts are changing
And since the specific topic at hand is the newsletter, what the
owner needs to do is not only define the customers, but also to define
them relative to the newsletter. What percent of new customers
respond now, and over time? What percent of "old
customers" respond, now, and over time? What percent of
"best customers" respond, now, and over time? Knowing
these numbers would almost certainly help the owner understand why
response to the newsletter is falling overall. The owner
resolves to address this situation immediately by digging into the
data. Yes folks, the inevitable Drilling Down...
Next month, we'll follow the owner of IMissAsia.com down the path
of defining, measuring, and tracking customer types. Which group
is responsible for the decline in newsletter response, and what can be
done about it? Only the data knows for sure...
To read the next installment of Recency: The Web Retailing Example,
I can teach you and your staff the basics of high ROI
customer marketing using your business model and
customer data, and without using a lot of fancy software. Not ready for the expense and resource drain of CRM?
Get CRM benefits using existing resources by scheduling
Questions from Fellow Drillers
If you still don't know what RFM is and how it can be used to drive increased profitability in just about any business, read this:
**Note to readers**: the promotion being discussed below is for an
offline retailer using direct mail. If you're not hip to control
and test groups or halo effects, you might want to read this
Q: I've really enjoyed your book and software.
I'm a business consultant with the University of (deleted). I'm
in the process of developing a Continuing Education class (2 1/2 hour
program) on Customer Database Management. I plan to reference
your book extensively and show off your software. I will be sure
to provide attendees your web address where they can order your
A: Well, I'm thrilled you like it and thanks for the
Q: One of the areas I'm working on now, and would appreciate
your input (perhaps a future newsletter topic), is what is the best
way to organize data (target / control lists) for calculating ROI on
A: Typically this is done with what is called a
"promotional history table," which can either be part of the
customer record or a unique table keyed by customer ID. Each
promotion has an "ID" and if the customer is selected for
the promotion, the promotion ID is placed in the table with customer
ID. So you end up with (to use a spreadsheet analogy) for each
row with a customer ID a list in the columns of the row flagging
promotions they have been in. This approach, of course, can
create a non-symmetric table and can lead to issues down the line.
Another way to do it, which is more difficult to execute but
sometimes preferred depending on what you are doing, is to have (to
use a spreadsheet analogy) each row represent a customer and each
column represent a promotion. If the customer was in the
promotion, the intersection of customer and promotion is
"Y." If the customer was not in the promotion,
the intersection of customer and promotion is "N."
This keeps the table
symmetric and can make querying easier.
Q: I am trying to put together one promotion (for the client I
mentioned above) and realized they have to track their promotion
target list and control list. If they begin doing a promotion
every month, or every week, that's going to grow into a large number
A: Yes, if you use lists. I can't imagine weekly
promotions for this kind of biz....unless they are going to different
people each time... Using the table approach described above
might be easier.
Q: Also, how do you handle a case where a customer was
targeted for a promotion, does not respond, and it's time to do
another promotion? Wouldn't that customer be included in the
next promotion (assuming that you have not given up, or written off
the customer as gone), and if so, how would you handle the ROI
calculation(s) if that customer responded to the next promotion?
A: As soon as you mail to that customer again you have
"poisoned" the ROI measurement of the original mailing; you
would have to cut off your measurement period before the second
mailing if you wanted ROI on the original mailing. But I
doubt you will ever be able to measure the true effectiveness of the
promotion with such short (weekly) time frames, because you cut-off
any of the halo effects the
promotion may have generated by stomping all over it with the results
of the next promotion.
Now, you might not care about that, and I don't really know what
the objective of the campaign is. But if the objective is to
maximize ROI, you won't be able to measure it if you poison the
control group so quickly. Wait 30 days if you are just looking to
measure the ROI of the promotion itself; wait 90 days if you are
looking at the dynamics and ROI of customer retention campaigns
If your plan is to sequence mailings, that is,
"the promotion" is actually 4 successive weekly mailers, and
you are measuring that against control, then you can mail every week
and mail to whoever you want, as long as it isn't people in
control. You can set up a "decision tree" if you want
that says "if they don't respond to #1, send #2 the next week; if
they do respond to #1, skip a week and send #3 in week 3, that kind of
Realize this though:
1. You will not be able to measure the effect of any one
mailer or decision tree sequence, only the effect of the entire
promotion. So you end up with a lot of work and you don't know
what was effective.
2. You can measure the effectiveness of an individual
piece or sequence, you just have to set up control for it at each new
branch or step, for example:
Mailer #1: Initial mailer, total = 1000
Test: 900 are mailed
Control: 100 are held back
You get results of: Non-responders: 500 Responders: 400
Mailer #2: Re-mail non-responders = 500
Test: 450 are mailed
Control: 50 are held back
Mailer #3: Re-mail responders = 400
Test: 360 are mailed
Control: 40 are held back
So for a simple 2 step mailing, you have 3 control groups and
3 test groups, and you can measure the effectiveness not only of the
total campaign, but each piece of it. If you don't set up
control at each step, if the overall campaign is successful, you won't
know if it came from initial response or the re-mails of responders or
The ROI you asked about above would be the ROI of Mailer #2 - did
not respond to initial mailer and were mailed again. As long as
control is composed of other people who received the first mailing and
did not respond, you should be able to measure ROI very accurately for
If you are a consultant, agency, or software developer with clients needing action-oriented customer modeling or High ROI Customer Marketing program designs,
If you are in SEO and the client isn't converting the additional
visitors you generate, click here.
That's it for this month's edition of the Drilling Down Newsletter. If you like the newsletter, please forward it to a friend - why don't you do this now while you are thinking of it? Subscription instructions are at the top and bottom of the newsletter for their convenience when subscribing.
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 Defection right here.
'Til next time, keep Drilling Down!
- Jim Novo
Copyright 2002, 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
attribution, including live web site link and/or e-mail link. Please
tell me where & when the material will appear.