Conversion Rate Optimisation
What we've learned so far
Conversion work is the cheapest growth available to most businesses. Traffic costs money every month and the percentage that converts is fixed capital: improve it once and every future visit is worth more. The same spend produces more revenue, the payback period on acquisition shortens, and the improvement compounds rather than resetting when the campaign budget does.
It needs doing rigorously because the field is full of confident nonsense. Most of the gains we have found were not where the industry writes about them. Button colour and headline tests produce small, noisy results. The larger wins came from removing a step, fixing a form that rejected valid postcodes, making shipping cost visible before the final screen, and repairing the mobile checkout nobody on the team uses. We start with the funnel and session recordings rather than a hypothesis list. The other discipline is statistical. A site with moderate volume can take weeks to resolve a five percent effect, and running a test for three days produces a number that looks like a result and is not one. We size a test before building it, and we say plainly when a site does not have the traffic to test at all, because on low volume the honest answer is research and qualified judgement rather than an experiment.
Generating variants is now trivial, which makes the discipline more valuable, not less. Twenty versions of a page can be produced in an hour, and without a way to tell which is better you have simply multiplied the guesswork. What we bring is knowing what to test, what not to bother testing, when the result is real, and when the right answer is to change the offer rather than the interface.
What this can involve
Experiment Backlog Management
Behavioural Analytics
UX Improvement Sprints
Funnel Analysis
A/B Test Programmes
Conversion Experiments
Launch Landing Pages
Product Page Copy
Lead Capture and Forms
Landing Page Optimisation
Shopify Conversion Optimisation
Checkout Optimisation
Ecommerce Store Audit
Session Recording and Heatmap Analysis
Form Design
How we work
Understand the main funnels
Analytics across the whole path from arrival to purchase, to locate the step losing the most value. It is often earlier than the step where the drop-off shows.
Work out why people leave there
Session recordings, support tickets and a handful of user sessions. The number tells you where. Only watching tells you what to change.
Fix what is clearly broken first
Errors, dead ends and confusion that needs no experiment to justify removing. Testing obvious faults wastes weeks proving what everyone already knows.
Test the changes worth arguing about
Where the answer is genuinely unclear and the traffic supports it, we run the test properly and let it finish rather than calling it early.
Keep what worked and move on
Winners shipped, losers documented so nobody proposes them again next quarter, and the next leak identified.
Where this isn't the right fit
If traffic is low, testing will not give you an answer. Below a few hundred conversions a month most tests never reach significance, and running them anyway produces confident decisions based on noise. At that volume, fix what is obviously broken and spend the rest on getting more of the right visitors.
If the traffic is the wrong traffic, no amount of page work will fix it. A checkout cannot rescue a visitor who was never going to buy. And if the product or the pricing is the problem, conversion work will keep finding the same ceiling every quarter. That is worth naming early rather than billing you to rediscover it.
