Testing Tactics Blog

So you heard about Taguchi testing, eh? You heard some big claims, too? People saying they increase marketing response by hundreds… thousands of percent… talking like a crazy advertising scientist?

Is it all lies or is it the truth?

I’ll break it down for you here so you can understand what Taguchi testing is — and what all the fuss is about.

First, the history — Dr. Gen’ichi Taguchi was an engineer in Japan. He was looking for a way to design experiments in manufacturing that created the best quality of products — and would consistently do so, over and over again.

He came up with some very advanced statistics that you could use to test a number of different variables to see how they worked together. The cool thing here is that with his statistics you could figure out how thousands of tests would work just by running a few.

Fast forward… one of Dr. Taguchi’s students, Dr. James Kowalick was working in manufacturing in the United States. One day he was walking down the hallway when he overheard the marketing department at his company talking about the same quality problems Dr. Taguchi had solved in manufacturing.

Dr. Kowalick offered his help, and the use of Taguchi testing in marketing and advertising was born.

Since then, Dr. Kowalick has worked as a high-level consultant in applying the Taguchi method to advertising, and has only taken on a couple students to teach his methods to, including David Bullock. Much of what this article is based on is David Bullock’s approach.

Now, the method — the Taguchi method relies heavily on statistics, but to run a Taguchi test, most marketers can get by with not knowing the statistics part. There are tools that will do all the fancy calculations for you.

The important part is understanding what to plug into your test to get the best result.

You see, Taguchi tests are based on plugging variables into “arrays.” These Taguchi arrays are based on Taguchi’s original formulas plus new formulas Dr. Kowalick came up with to specifically apply to advertising. The arrays work as road maps for your tests.

The arrays tell you how many different versions of each variable you can put in, then the array mixes them up to tell you what to test.

For example, one common array lets you test 6 versions of one variable, and 3 versions of six different variables.

This means you could test six headlines, three pre-headlines, three sub-headlines, three pictures, three call-to-actions, three guarantees, and three order buttons.

Altogether that would create 4374 combinations, but Taguchi narrows that down to only 18 tests necessary to figure out what each of the 4374 would do.

When you run those tests and plug the results into your Taguchi statistics program, it kicks back which of the 4374 possible combinations would give you the best results.

Often times — with the right input — this will create increases in response rate of over 100%, and sometimes as high as 2300% or more.

The beauty of Taguchi testing and the Taguchi method is how fast it lets you test all the different combinations of variables. It gets you increases which would take years of consecutive testing with previous testing methods.

So… as for Taguchi testing and its ability to increase your profits through optimizing your marketing…

Is it all lies? No. Are there some people in the marketplace exaggerating claims? Probably. Is it worth learning more about? Definitely.

---

>>> Discover quick and easy ways to improve your marketing so you can make more sales while spending less. Listen in on a FREE 90-minute interview on the "7 Principles of Taguchi Marketing" with marketing testing and optimization expert David Bullock. Plus get a free e-mail course on applying what you learn in the call -- along with some extra free goodies.

Sign up here and we'll send you the download link right away:

First Name:
Email:

Click this button, then
immediately check your e-mail.

We hate spam too -- We won't send you junk and you can unsubscribe at any time... but we're pretty sure you won't want to!