Dan Kennedy: Test And Grow Rich

“Testing” is an ugly topic.

Why? Because testing variables in advertising, direct mail, phone scripts and sales presentations requires discipline, diligence and patience. To get it right, you can only test one variable at a time. This means that if you change a headline, you can’t change anything else. Plus you have to make sure all other variables remain the same, like the mailing day or a war breaking out that has everybody watching CNN day and night or another politician getting caught again with his drawers down or a hurricane hitting.

Frankly, most business people will just not go through the “detailitis” required to test – which is why it’s a very good idea to model proven promotions. And in some cases where you’re only going to use something once or twice or you’re dealing with a very small number, it’s just not worth testing; instead, you take your best shot. But let’s assume you’re working on something you intend to use over and over and over again in some significant quantity, so that it’s worth real effort to fine-tune it…

I have some tips for you: first of all, there’s non-testing testing – huh?

Well, I describe that in my book ‘The Ultimate Sales Letter’, where I talk about the steps to take with a finished sales letter before you actually mail it.

Second, there’s split testing, which is the fastest way to test and get to a reasonable conclusion. Let’s assume you have a postcard and you want to leave everything the same but test four different headlines, and you have 4,000 similar addresses to mail to. You do “nth name testing”; that means Headline #A goes to every 4th name, Headline #B to every 4th name, Headline #C to every 4th name, etc. So you evenly divide the list without bias among the headlines being tested. Some media (like Val-Pak or MoneyMailer) will let you split test within a single buy.

Third, there’s testing against a control. A “control” is a marketing strategy that already works well and you’re using it on a continuing basis – maybe it’s a series of letters you mail every month. You have been using it long enough you know what it produces. You have a “known” to measure against. Now you can start trying to improve that control, ideally one step or variable at a time.

If I’m trying to beat a control, here are the “hot” variables I’ll look at closely, to see if there’s room for improvement:

1. The offer
2. The guarantee(s)
3. The urgency of response
4. The big idea or big promise
5. The overcoming of skepticism i.e. credibility and believability
6. The style or tone of the writing itself
7. The look of the piece

By the way, little, very testable things DO sometimes make very big differences. Recently I showed an example in my newsletter of a guy who just added four rubber-stamped words to the outside of his envelope, and beat his control by 300%. I once brought a TV infomercial back from the dead by raising the price of the product.

Gary Halbert saved the Pearl Cream advertising by adding a particular bonus. In 1984, after attending my seminar, a dentist in Sacramento changed five words on his Val-Pak coupon and went from getting two or three new patients a month to 15 to 20. This is the sort of thing that makes direct-response advertising as frustrating as golf. (Remember when golf pro John Daly scored an 18 on a single hole? Ugh.)

Obviously, you can’t test if you can’t, won’t or don’t collect accurate data. You have to code every offer, and track where every ounce of business comes from. If you have employees who are lax about this, you must educate them about the importance, discipline them if they goof it up, and ultimately can ’em if they won’t do it right 100% of the time.

I confess that I fly by the seat of my pants in my business more than I should, but I can’t fire me, God knows there are days I should. Anyway, I can assure you: the clients I have with the best profits and incomes possess the best information about where their business comes from.

Let me switch gears and talk briefly about another aspect of “testing”. This is actually how all highly successful entrepreneurs view everything they do…as testing. They do NOT see things in the context of “success” or “failure” like ordinary people do, and as a result they do not become “de-motivated” like most people do. See, most people drain all the vitality, courage, optimism and git-up-n-go out of themselves by focusing on all the things they do that don’t work out well, as a compilation of failures.

But successful people understand the powerful impact of that negative reinforcement on their own self-image (somewhat akin to the impact of pouring a gallon of toxic waste into a pint of clear water, drinking the result, and wondering why the stomach backs up into the esophagus). Instead, they carefully organize the things they do into a series or sequence of experiments, testing options, and focusing on the ones they find that work. And they fully expect to go through any number of experiments that don’t pan out before walking away from the lab with a winner. This not only has practical relevance, it has profound psychological ramifications.

Just like a little tweak in thinking can make a big difference in the results of say, an ad or a flyer, a little tweak in thinking can make a giant difference in the life results experienced by an individual.