Six Months of Pricing Page Tests, What Actually Moved Revenue

Over six months we ran twelve distinct tests on a client's pricing page: anchor pricing order, annual-vs-monthly default toggle, feature comparison table layout, badge placement on the recommended plan, urgency copy, social proof placement, and a few others borrowed from the standard CRO playbook. Ten of the twelve produced no statistically significant change in revenue per visitor. The two that did weren't the ones anyone predicted.

What didn't move anything, despite the playbook

Urgency copy ('limited time offer') did nothing — this is B2B software with a considered sales cycle, not an impulse purchase, and the audience saw straight through it. Badge placement on the recommended plan did nothing measurable either; by the time someone's on a pricing page, they've usually already decided roughly which tier fits and a badge doesn't change that. Both are tactics that work well documented for e-commerce and simply don't transfer to a longer-consideration B2B purchase.

Half the CRO playbook is genuinely effective and the other half only works for the use case it was invented for. The expensive mistake is applying all of it without checking which half you're in.

What actually moved revenue

• Defaulting the annual/monthly toggle to annual (rather than letting monthly be the default view) increased annual plan selection meaningfully, since most visitors never touched the toggle at all.
• Rewriting the feature comparison table from a generic feature list to outcome-based language specific to the visitor's likely use case moved mid-tier conversion noticeably, because it answered 'is this plan actually for someone like me' better than a checkbox grid did.
• Everything else tested — pricing order, badges, urgency copy, social proof placement — showed no statistically significant revenue impact over the test duration.

If there's a generalizable lesson from six months of this, it's that the tests which worked were the ones that changed what information the visitor had to make a decision, not the ones that changed how the page looked or felt. Cosmetic and psychological nudges mostly didn't survive contact with a B2B audience that was already deep into evaluating the product on its merits.