The Taguchi Method
How it works
With Multivariate testing even with just five different elements on a web page tested for e.g. Headline, product image, buy now button, with just five different alternatives for each element you need to run 125 different combinations.
You would really need to see over 150 conversions for each combination to obtain a conclusive test result. The traffic volume generated from these results would be too great to handle for most applications.
To get round this obstacle testing needs to made smaller with an indicative subset of combinations and elements which would estimate the best theoretical combination if it was tested over and over this is called “Fractional factorial testing” and can be done using a method called Taguchi.
You can easily estimate the elements that work for your web pages with Taguchi as it will give you exact combinations to use.
Why we use Taguchi
- Small testing cycle and minimal traffic required. The testing only takes a couple of days or up to a month with varying combinations and elements.
- Time saving the Taguchi tests can be designed and run quickly using modular elements.
- Great Improvement using the best combinations for web pages your results will speak volumes.
- Easily identify the elements that increase conversion rates
- High impact elements tested with a great number of alternatives
- Marketing experience not required you don't need to be a statistician or hold an expert marketing position to design a Taguchi test cycle.
Taguchi facts
- The Taguchi method is flexible and advanced in its approach and can be used to approach multiple pages and sessions this has been demonstrated with success.
- The Taguchi method used in testing will return regular improvements to your results over and above 25% and has also exceeded 90%.
Taguchi history
Taguchi is not a new concept and has been around for 50 years, used with great effect in the manufacturing industry and car production but now applied to web applications and direct mailing companies.