Conversion Rate Optimisation
A/B Testing
A test that stops early or measures the wrong metric is worse than no test. The discipline is in the design, not the tool.
- Engagement
- Monthly retainer, with a minimum testing window.
- Outcomes tracked
- Lift per testTest velocityWin rate
Included
What this covers.
Delivered by the same team that builds your site and systems, so the work is not handed between agencies.
- Hypothesis backlog with expected impact
- Sample size calculation before launch
- Test implementation and QA
- Single-metric decision rules
- Segment analysis for insight
- Winner rollout and documentation
In detail
Who it suits, and how it works.
Best fit
Businesses with enough traffic to reach significance, or with high-value pages where small lifts are worth real money.
Design decides the result
Sample size is computed before launch and the decision rule is agreed in advance. Stopping when a variant looks good is how false wins get shipped.
Learning compounds
Even a flat test teaches something about the audience. Documented hypotheses turn testing from a series of bets into a body of knowledge.
Questions
Before you commit.
- How many tests should we run per month?
- One to two well-designed tests beat five rushed ones. Each needs time to reach significance.
- What if a test is inconclusive?
- That is a legitimate result. We document it and move on rather than declaring a winner on a flat result.
More within Conversion Rate Optimisation
View Conversion Rate OptimisationNext step
Tell us what you are trying to fix.
A short conversation is usually enough to know whether we are the right fit. If we are not, we will say so.