Morgan Dutemple
All expertise areas

CRO: convert better before buying more traffic

A good share of the projects I manage are e-commerce sites or sites with a strong conversion logic: checkout funnels, product pages, lead generation. A funnel that converts poorly is as much a scoping problem as a marketing one, and changing a journey also means changing habits. I built a page comparator that judges each page type against its own criteria.

Why more traffic is not enough

The most common reflex when online revenue disappoints is to buy more traffic. But a funnel that converts poorly stays a funnel that converts poorly, however many visitors you send into it: pouring more water into a leaking bucket does not plug the hole.

CRO starts by finding where and why visitors drop off, not by changing the colour of a button. Real observation (session recordings, heatmaps, a journey tested by hand) already reveals most friction points without waiting for the volume an A/B test needs. And a homepage, a product page and a checkout are not judged against the same criteria.

My method

I do not treat CRO as a separate marketing topic: it is part of scoping, like the rest of the project.

  1. Observe the real journey

    Session recordings, heatmaps, a journey tested by hand: where visitors drop off.

  2. Prioritise by impact and effort

    What really blocks conversion before what merely looks important.

  3. Test with an explicit hypothesis

    An A/B test when volume allows, a user test otherwise; never a change without measurement.

  4. Build it into scoping

    The fix goes into the product backlog, not into a separate list that gets forgotten.

In a web agency: lead generation tooled with AI

On a web project run in an agency, AI also served lead generation: producing variants of landing pages and wording faster than by hand. It is quick at proposing options, which suits CRO well: CRO is about comparing them.

The decision is not its call. Each variant starts from a written hypothesis, is judged on the real journey and goes through the same review as the rest of the project. Without a hypothesis, more variants only means more noise.

Frequently asked questions

What is CRO (conversion rate optimisation)?

CRO covers the methods that increase the share of visitors who complete the expected action on a site (purchase, form, contact request), without necessarily increasing traffic. It relies on analysing real behaviour, prioritising friction points and running controlled tests before rolling out a change.

Does CRO require high traffic volume?

A reliable A/B test needs a minimum volume, but CRO is not limited to testing. Direct observation (session recordings, heatmaps, a journey tested by hand) already finds a large share of friction points, whatever the traffic.

Which page should I start with?

The one where drop-off is highest in your data, often the basket or the payment step. Check it in your analytics before choosing: intuition rarely picks the right page.