The problem
Most A/B tests are too small to detect the effect they're looking for, and stop the moment a result looks good. Both lead to changes that don't actually help.
Services
We run research-led testing programmes on your site and funnels, designed to detect real effects and read out without wishful thinking.
Most A/B tests are too small to detect the effect they're looking for, and stop the moment a result looks good. Both lead to changes that don't actually help.
Analytics, user research and heuristic review.
Ranked by expected impact and evidence.
Sample size and duration worked out before launch.
Tests built and checked across devices.
The effect, its range and how sure we are.
What to ship, and what to learn next.
Every readout shows the range of likely effects and the chance the change really helps.
Research finds the problems worth solving. Every test is sized before launch, built and checked, then read out once at its planned end, with the likely range of the effect, not just a pass or fail.
The decision and the constraints.
A fixed-scope look at where you are.
Agreed scope, regular check-ins.
Documentation and training.
FAQ
We work in three-month blocks, with a review at the end of each.
We work in the platform you already use, including VWO, Optimizely and Adobe Target, among others, as well as server-side and in-house testing frameworks. If you don't have one yet, we'll help you choose based on your traffic, stack and budget.
Thirty minutes, no slides.