Split Run Testing

The "split run" is a testing device long employed in mail order advertising. You have two ads. Which is more effective? You get the answer with a split run test. In addition to testing the effectiveness of an entire ad, headlines, photos, color vs. black and white, etc. can all be tested.

The mechanics for a split run test are simple. The advertiser prepares two ads of identical size and gives each a separate key code on its order coupon. When the coupons (orders) come in, the numbers are read and the winner declared.

The "perfect A-B split" is made possible thanks to the mechanics of the printing press. Some (not all) publishers of newspapers and magazines are able to print in such a way that, as the pages come off the press and are bound, alternate "books" include first the "A," then the "B," then the "A" ad again, etc. Thus a "perfect" split as the publication goes into distribution.

But now we're on the internet. How do we test ads and advertising concepts? It isn't easy and the attempts we make produce much less precise results than a "perfect" A-B split. Yet there are times when we still want to give the concept a try.

I have a number of fragrances that I have created. By myself. My own formulas. And I sell them on the internet. I have been developing a website, FrankBush.com and that website has produced some sales.

But, since I'm pretty much unknown as a perfumer outside of my own blogs, I need to stand on my head to get attention -- like the owner of a used car lot -- so to give the FrankBush.com website some "personality" I have tried to make it a bit colorful and not exactly what most perfume buyers would expect.

FrankBush.com website soldiers along.

But could I do better? Would a more "serious" approach to selling my own perfume outdo the seemingly frivolous approach of FrankBush.com? Thanks to the low cost of setting up a new website, I decided I would give it a try. Thus I now have a second website selling my own perfume and men's fragrances, a "serious" site, PGLightyears.com.

This split test will never have the accuracy of a traditional mail order perfect A-B split. But, as the new site acquires visitors (this will take time), I will be able to see whether I get more sample requests from the one or from the other. This information will help me develop new dvertising pages that sell more effectively.

Ultimately, it's the perfume itself that will make or break the business. But without a "following" you can't expect sales and I want to develop effective communications to build that following. I'm sure that you want this too.