Build
A/B Testing Your Pre-Sell Pages — What to Test First and What Doesn't Matter
You've got a pre-sell page. It's converting. But is it converting as well as it could? The answer is almost certainly no — and the testing order matters more than the tests themselves. Most operators waste their first 90 days testing things that move the needle by 2% when they should be testing things that move it by 40%.
October 14, 2026 · 9 min read
There’s a scene in Moneyball where Billy Beane sits across from his scouts and they’re arguing about a player’s jaw structure. His girlfriend. Whether he “looks like a ballplayer.” The room is full of confident men debating the wrong variables with absolute certainty.
Meanwhile, the kid in the corner with the laptop has the numbers that actually predict wins.
I think about that scene every time I see an operator agonize over their button color. Or their font pairing. Or whether “Submit” converts better than “Get Access” on a page that only gets 200 visitors a month.
You’re arguing about jaw structure. The on-base percentage is sitting right there in the data, and nobody’s looking at it.
Pre-sell pages sit between traffic and merchant — they’re the toll position made visible. And like any piece of infrastructure, they can be optimized. But the testing order matters enormously. Some elements can double your conversion. Others will waste three months of traffic on a 1.5% lift that’s probably statistical noise anyway.
Here’s the hierarchy. It’s not a suggestion — it’s backed by the conversion data I’ve watched accumulate across dozens of pre-sell pages over two years.
Headline > CTA copy and placement > Page length and structure > Everything else.
That last category — “everything else” — includes colors, fonts, button shapes, hero images, and the twelve other things you’re tempted to test first because they’re fun. They are, at best, 2-5% variations. The top of the hierarchy routinely produces 20-60% swings.
This article continues for paid subscribers with the full testing hierarchy (what to test and in what order), sample size requirements for statistical significance, the three mistakes that poison your test data, and a 90-day testing calendar.
Paid Content
This article continues for subscribers.
The full implementation — the steps, the templates, the math — is delivered via the newsletter.
Unsubscribe in one click. Useful operator material only. Promotions, sponsors, and partner links will be disclosed.
Build
More on building toll infrastructure?
This article covered one piece of the build dimension. The newsletter delivers a new install pattern every week.
Unsubscribe in one click. Useful operator material only. Promotions, sponsors, and partner links will be disclosed.