You want to structure your A-B test to zero in on the one variable that will provide your greatest return. Here is a list of some variables that you can test:

Price A versus Price B

Radio ad A versus Radio ad B

Test headlines

Test newspaper ads

Test left-side of the page versus right-side of the page

Test different direct mailings

Test different response forms

Test different sales pitches

Test different offers

Test different packages

Test different upgrades

Test different back-ends

Test different credit policies

Test different guarantees

Test different incentives

Let’s use an example. Pricing. Which price is proven more attractive to a prospect when a $10.00 item is tested as the price for two of the same item? $9.00 for one and $1.00 for the other. 50% off both items; $10.00 for one and one is free. We’re basically giving them several options.

“Buy one and get one free” has been shown to outpull the other two approaches. Even though “Buy one for $9.00, the other for $1.00” and “Buy 50% off both items” are really the same thing. You find this out by testing. These are different offers.

If you’re creating a direct mail campaign, you create two different mailing pieces offering your product two different ways. “50% off both items” or “Buy one, get one free.” You send them out and tabulate the response to determine the best price to use and set your marketing campaign with that chosen price.

You don’t want to stop there. Send out one with one headline and one with a different headline. You keep testing. Test one piece at a time because if you change a headline, a price and a guarantee, and your response goes up or down, you won’t know why.