One of the biggest keys to the ongoing success of any marketing campaign is to understand that it's not perfect. Results can almost always be improved; but figuring out how to get those improvements, that's done by testing.

Too many marketing efforts have gone down the tubes because marketers "went with their gut" on a campaign, believing they knew best what the customer would want. Remember that "New & Improved" ... sometimes isn't.

In truth, it is your customers that will tell you what's working and what's not; you simply have to give them a chance, through logical, controlled testing programs.

You can test practically anything. Price points; merchandise mix; website design; advertising message; and so on. All the elements that go into a buying decision can and should be tested.

Here are some key guidelines to effective marketing testing:

*Test against a control. You learned back in science class that you've got to have a standard that your test runs against, to see if you're actually making any improvements. Your "control" is your current product, or sales copy, or whatever. Your current performance is the baseline you want to test against. Run your test against your control, head-to-head, at the same time.

*Test the big stuff. As a general rule, you want to invest the time, energy and money into testing the things that can make a big difference. Example: Rather than testing a change in the color scheme of your website, test a complete redesign.

*Test the right stuff. The right stuff to test is stuff that's scalable if it works - meaning, it could be rolled out and become your control. Example: If you tested offering a specific item as a premium, you'd want to be sure that item was available on an ongoing basis, should the test work.

*Test one thing at a time. Another lesson you learned in science. Limit your variables, or you won't know what made a difference. Example: Don't test a redesign of your website AND a premium offer at the same time. If results improve, you won't know which made the difference.

*Tracking is the key. Be sure you've got a solid tracking process in place, so you can correctly analyze your test results.

*Test continuously. All tests should have a specific start and stop date - but you should never stop testing. Once one test is complete and you've analyzed the data, start testing something else.

Remember, no matter how well you're doing, you can almost always do better. So test your way to even greater success!

(iamsbm)