When you have a control package that’s performing near or on budget, you may not want to risk testing a completely different approach. You should, however, test “tweaks” to the control package that have the potential to generate small increases in response/profitability. Theoretically, a series of successful tweaks will add up to a major improvement in your program.
There is, however, a danger in tweaking. It results from misunderstanding the difference between a tweak and a substantive (and possibly damaging) change to the control package.
Here’s an example from the world of publishing. The control package is comprised of a promotional outer envelope; a four-page, 8½˝ x 11˝ letter; a four-page, full-color brochure; a lift letter; a response form with offer labels; and a BRE.
Outer envelope (OE): The control has a teaser on the front that says “What if _______ is good for you?” If you change that to read “Is ______ good for you?” that’s definitely a tweak. If you move copy about the free issue from the back of the envelope to the front, that’s a tweak. If you change the background color of the envelope, that’s also a tweak. You are not dancing tweak to tweak if you take the question off the OE and use assumptive copy—e.g., “A free issue of XYZ Magazine is awaiting your confirmation.”
Letter: After the Johnson box, the control letter answers the question posed on the OE, then provides a series of bullet points covering benefits, specific content of the magazine and subscriber benefits. It also includes testimonials. If you change some of the bullets relating to content (to update them) and the testimonials (you believe you have stronger ones), these would be considered tweaks. If you change the opening and don’t answer the question on the envelope, that is much more than a tweak. What if you change the Johnson box completely, but leave everything else in the control alone? I would contend that’s a tweak as well, but it’s certainly open to debate.