A/B testing benchmarks

What actually wins, from thousands of real A/B tests run on Mida and concluded between August 2023 and August 2026. A win means the variant beat its control with a statistically significant result.

By what you change

Bars are win rate; the line through each is the range the true rate is likely to sit in — wider where fewer tests ran. The leader by rate rests on the fewest tests, so read the ranges, not the order. Faded bars are based on fewer than ten wins, and an arrow means the range continues past the axis; hover any row for its exact figures.

change type	win rate	tests
CTA colour	16%	32 tests
Layout	15%	250+
Form	12%	50+
CTA copy	12%	250+
Hero image	12%	100+
Price framing	11%	100+
Styling	11%	1,000+
Split URL	10%	1,000+
Headline	8%	500+
Social proof	8%	100+
Body copy	6%	250+

These are not a ranking. How often a test wins depends partly on what it measured — proxy goals clear the bar 13% of the time against 9% for business outcomes, and the mix varies by category. Compare like for like, or read the methodology first.

By industry

industry	win rate	variants
E-commerce	10%	1840
SaaS / Software	9%	726
Professional Services	12%	496
Finance & Insurance	12%	428
Education	9%	422
Health & Wellness	12%	361
Media & Publishing	12%	199
Marketplace	8%	191
Travel & Hospitality	8%	131
Healthcare	6%	110
Nonprofit	10%	40

Industry comes from the verified domain the Mida pixel fired on, covering 97.8% of tested variants — not from a self-reported field, which disagrees with it 65% of the time.

Read the method before the numbers

Most tests do not win, and most of what gets tested never moves anything. We exclude tests that never got enough traffic to say anything either way, because a flat result on too little traffic is not evidence a change failed. The methodology page states exactly what we exclude and what these numbers cannot tell you.

Who ran these tests

This is every Mida account that ran a readable test — in-house marketers, founders, product teams, and agencies working on client sites. Nothing here is filtered by who ran the experiment or how experienced they are.

Low win rates are normal in experimentation, including at the top end. Microsoft's experimentation team, reporting on its own platform, found that only about one third of ideas improve the metric they were designed to improve — and that roughly another third actively hurt it. That is a dedicated experimentation organisation with research, prioritisation and review behind every test.

A mixed population like this one runs below that. The gap is roughly what disciplined practice buys you: ideas grounded in research rather than opinion, one variable at a time, and tests built so the result can actually be read.

So treat these as a general base rate, not a target. If you would rather not close that gap the slow way, find a CRO agency in the United States, the United Kingdom, the Netherlands or Australia — or browse every region.

A win is a variant that beat its control on that test's primary goal with a statistically significant result. Tests that never got enough traffic to say anything either way are excluded. The methodology has the full detail, including what these numbers cannot tell you.