Science in direct marketing

There's an earlier post on hypothesis testing.
http://imaginendless.blogspot.com/2010/08/hypothesis-testing.html

The example I'll use is what was used for me when I worked at a direct
marketing business.

Lets say you want to send out a mail shot and are giving a choice of
blue or yellow envelopes. Most people would likely chose to use the
yellow envelope because they might think people will be more likely to
open brightly colour envelopes.

Evidence-based, or direct, marketing doesn't think that way. Both blue
and yellow envelopes are sent out with the same letter to the same
target group as an experiment. Only one thing is changed - the colour of
the envelope - in the experiment. The result can then be used in the
next mail shot.

Direct marketing companies take this science to a new level. The
subtlest differences are tested.

It's important to consider the options though. There are many choices of
envelope and it may be surprising what people are most likely to open.
It's important to stay as open-minded as possible when choosing the
variables.

An amusing blunder at a an bank shows what I mean. In the 1980s a bank
with a large number of high value clients (millionaires) tried using
mail merge to customise each letter from the information on their client
database. In this era there wasn't the convenience of Microsoft Office's
mail merge wizard. It was a bespoke application written from scratch.

In the "Dear.." field the programmers used the test name "Rich bastard".
You can see where this is going. By accident the programmers forgot to
take it out of the code and no one checked so the letter went out to
their richest clients calling them Rich Bastard.

This letter got an exceptionally high response rate. The responses
turned into sales too. The investment opportunity that was offered was a
good one and this may have been a factor. It was the 1980s though. Harry
Enfield was ribbing the rich with his Loads of money character. The
letter may have been seen as creative marketing and it definitely worked.

Creativity is important in science. The willingness to try something
very different in an experiment can lead to new avenues. This applies to
the science of direct marketing too.