CRO and A/B Testing

Fewer tests, on the pages that matter.

Webtonic Uslege SGC Ready Set Startup Rapidgrade Qazverse Qazis Toolkit Proal Jovana Fizio Edenhauser Cinnamon Hearst Dom Učenika Webtonic Uslege SGC Ready Set Startup Rapidgrade Qazverse Qazis Toolkit Proal Jovana Fizio Edenhauser Cinnamon Hearst Dom Učenika

Most B2B sites do not have the traffic for a new test every week. We pick the busiest pages, test bigger changes there, and run a QA step before any result is believed, so the wins you ship are real.

Book a strategy call See how we work →
How we work

A testing programme sized to your traffic.

The plan starts from how many conversions you get, not from a list of ideas. That number decides how many tests are worth running.

01

Research

Find where people drop.

Analytics, form data and session recordings on the pages that carry the most traffic, to find the step where buyers give up.

  • Funnel analysis
  • Form drop-off
  • Session recordings
02

Test plan

Ranked by traffic and impact.

A short list of tests, each with the sample size it needs before it starts, so nothing runs for months without a chance of a result.

  • Sample size per test
  • Bigger changes first
  • One primary metric
03

QA

Checked before it is believed.

Every variant checked on real devices, tracking verified, and traffic split checked, because a broken variant looks exactly like a losing one.

  • Variant QA
  • Tracking checks
  • Traffic split checks
04

Ship and document

Winners shipped, losers learned from.

Winning variants built into the site properly, and every result written up so the next test starts from what you already know.

  • Winners built in
  • Results log
  • Next test chosen
From the blog

Win rates, tools and the traffic you need before a test is worth running.

All articles
FAQ

Questions we get early.

Everything you want to know before you reach out.

It depends on conversions, not visits: each variant needs enough of them to show a real difference. When there are too few, we run fewer, bigger tests, or make the change without a test and measure before and after.

Fewer than most people think. Optimizely's analysis of 173,000 experiments found 12% win on their primary metric. Our A/B testing statistics collect the rest of the published data.

We quote per project, based on how many pages are in scope and how much traffic they get.

Whichever fits your traffic and budget. We recommend one after seeing your numbers, and we can work in the tool you already pay for.

Test less. Learn more.

Thirty minutes. Tell us your monthly traffic and conversions, and we will tell you how many tests are worth running and where we would start.

Book a strategy call
Veljko Nikolic Mihajlo Stojanovic
You talk to the founders, not a sales rep