 |
|
Re: Is personalization really personal?
PAULINE CHIARELLI WROTE:
>Mark Montgomery mentions the high cost of personalisation
>and while I agree it's expensive and difficult, particularly
>for mass mailing, in my view it is, at least in part,
>essential. Most marketers fall for the "Some's good, More's
>better" approach and start standardising and cutting
>corners, treating customers as numbers.
In addition to web marketing, another hot area for
personalization is direct marketing. Here I am going to use
direct marketing as an example to illustrate what
personalization really means.
Suppose you have a well-maintained database with millions of
customer names and their demographic information and
purchase history information. Now you want to make your
direct marketing more personal (As a matter of fact, what
you really want is to increase your targeting accuracy and
to reduce your marketing cost. You don't care much about
your customers' personal experiences.) How do you reach your
goal?
Here you must introduce statistics though someone does not
like it and claims it is outdated.
Although there are numerous names for this targeting
process, the basic rule is to find the right people who have
higher probabilities of purchasing your products/services.
Although there are many statistical models that can help you
reach this goal, the most useful and the simplest model is
Logistic Regression. Other models just have marginal
improvement but need much more extra efforts.
Now the assumption about your customers plays a critical
role in your statistical analysis. Do your customers have
the same preferences? Do they have the same response
patterns? If you follow the one to one marketing assumption,
you have to treat each customer differently (because you
assume they are different.) In other words, you need to
estimate each customer's response pattern individually. (You
can not assume they are different but estimate their
response pattern at the aggregate level -- to assume they
are the same.) Now you need to answer the following
questions to make sure you are on the right track:
1. Are your customers' response patterns really different
from each other? How do you know that?
2. What portion of your customers has enough data points for
such individual estimation?
3. How costly is such individual estimation? What is the
marginal gain for such individual estimation?
I bet no one to one marketing advocator so far tries to
answer these questions.
Such direct marketing targeting estimation is not even at
the segment level but at the aggregate level. However it
might still give your customers "personal experiences".
For any personalization mechanism, you have to use
statistics to make predictions. If you use statistics in
such predictions, you have to estimate the algorithms at the
aggregate (or segment) level. Therefore, personalization
never assumes your customers are different from each other.
I have been doing such "personalization" for decades. Only
recently have I realized that I have been doing something
very popular on the Internet. But I am reluctant to call it
personalization because I never assume anything personal.
Regards,
Tim Lee
Editor -- Journal of Web Marketing Research
http://www.WebCMO.com
Innovations in web marketing & strategies
Participate in "Web Design Strategy" survey to get your EXCLUSIVE report
http://www.WebCMO.com/mis/survey.htm
========================================================================
------------------------------------------------------------------------
This week's Online Advertising Discussion List sponsors:
Flycast, Acceleration, and AdFlight
THE WORLD'S ONLY RESPONSE-ORIENTED ONLINE AD NETWORK.
With behavioral targeting and expert Media Consultants, we'll help
you reach 41% of the Internet. Visit http://www.flycast.com/oa2.
---
###100% GUARANTEED### Pay-Per-Click and Pay-Per-Download Advertising
All of your money back if you are not completely satisfied!
-->> http://www.clicksales.com/ and http://www.downloadsales.com/ <<--
---
AdFlight (formerly Immerce) - The Internet Advertising Exchange: Your
high choice, low price source for Internet ads! WEEKLY AUCTIONS,
DAILY DEALS! Go now to http://www.adflight.com/default.asp?source=oad
------------------------------------------------------------------------
========================================================================
Received on Tue Aug 10 1999 - 14:13:02 CDT
HOW TO JOIN THE ONLINE ADVERTISING DISCUSSION LIST
|
With an archive of more than 14,000 postings, since 1996 the
Online Advertising Discussion List has been the Internet's leading forum focused on professional discussion
of online advertising and online media buying and selling strategies, results, studies, tools, and media
coverage. If you wish to join the discussion list, please use this link to sign up on the home page of the Online Advertising Discussion List. |
|
|
Online Advertising Industry Leaders:
Clicksor
Local SEO with Video
AdJungle
Houston Web Design
The Laredo Group
Pay As You Go Advertising
Add your company...





|