In the old days, before marketers began to mistake technology for knowledge and sophistication, most companies relied on a variety of tried-and-true tactics to make the most of their marketing budgets. One of the most effective tools they used was called the split test.

Essentially, what they would do is make two different offers to two identical groups. Suppose they had a mailing list with 50,000 names. They would take 1000 names from that list, divide them into two groups, and send them offers that differed in content or approach.

For example, one offer might carry “buy one, get one free” creative, while the other proclaimed “half off.” Once the responses came in, they would analyze the results to see which did better, and then send that better-performing offer to the rest of the list. (Plus, they could project the results they received across the larger list. If 25 people of the original 500 responded, they could assume that a mailing to 50,000 would produce 2,500 orders — and they’d usually be right.

You can use split tests to compare creative, offers, format, design — pretty much anything you want. You can also do it in email, advertising, and other channels. Just be sure that you test only one variable at a time, or you’ll never be sure which one actually made the difference.