A/B Testing for Pricing
Curious if anyone has done testing with your prices, pricing display page / table, or even feature packaging to see what increases overall revenue and conversion.
If you've done this, could you share:
-what tools did you use?
-what were your results?
-was it worth doing this A/B testing?
Thanks!
On "was it worth it" - we run A/B tests for clients, so instead of guessing I pulled our own numbers. Across 2,288 tests, price-display tests won about 48% of the time. Copy and offer changes win more often (~60%), layout and trust-signal tweaks less (~42%). So pricing's roughly a coin flip, but when it wins it moves real money, so it's worth doing if you size the upside for that hit rate instead of expecting every test to land.
On the underpowered-variants point above - fair, and worth adding that most winners never clear a strict significance bar anyway. About 19% of our tests hit significance; ~50% just finished ahead on the primary metric. We shipped plenty of the ahead-but-not-significant ones when the read was clear and the downside was low. For pricing especially, watch revenue per visitor, not just conversion - a variant can lift CVR and quietly drop revenue.
On tools, anything that lets you split results by revenue works; the metric matters more than the tool.
When testing pricing, the biggest trap is underpowered variants. As a guardrail, I avoid reading results below ~200 qualified sessions per variant and keep cohorts consistent (new visitors only, fixed window). What sample-size rule do you use before calling a winner?