Most websites spend money (or time) bringing more people to a page that converts terribly. It is like pouring more water into a funnel without fixing the hole. Conversion optimisation — CRO — does the opposite: with the same traffic, more of the actions that matter.
First: what you measure today
You cannot improve what you do not measure. Before touching anything you need to know how many visits come in, what they do and where they leave. GA4 is your starting point: sessions per page, interaction events, abandonment and goals. If you do not know your current conversion rate, the first task is to calculate it.
- Conversion rate = actions that matter / visits. Without that number, everything else is guessing.
- Events: call, form, WhatsApp, primary button. Each one counts an action.
- Funnel: how many go from step 1 to step 2 to step 3. The hole appears there.
The three highest impact fixes
- A clear pitch in 5 seconds — the first screen must say what you sell, to whom and what to do. If they have to guess, they leave.
- One goal per page — two buttons with the same weight are zero buttons. Pick the action and point everything at it.
- Friction out of the way — short forms, fast load, copy that does not scroll forever on mobile. Every piece of friction you remove adds conversion.
How to run experiments that do not lie
Changing something and seeing if it worked requires comparing like for like: version A against version B, with the same traffic, the same time and the same criterion. A well-run experiment changes one variable, measures it and discards it if it does not win.
- One variable at a time: changing two things at once does not tell you which one worked.
- Enough size: decide beforehand with how many visits you can trust the result.
- Result on screen: the experiment is published with its number, not a feeling.
The example that explains it all
Imagine a business with 1,000 visits a month converting 1%: that is 10 clients. Bringing double the traffic (2,000 visits) without changing anything gives you 20 clients, at the price of doubling what you spend on acquisition. Meanwhile, raising conversion from 1% to 2% with the same traffic also gives you 20 clients, but without spending a euro more on traffic. Both paths land in the same place: one costs, the other makes what you already pay work harder.
And here is the nuance: doubling traffic is not free, not even once. Every new visitor costs acquisition, and cheap traffic today becomes more expensive tomorrow as ads compete. Conversion, on the other hand, is bought once and keeps paying off with every future visitor.
Where to look in GA4
The analysis session is not a place for opinions: it is the roadmap. GA4 gives you four views that order the work: the traffic line (where people come from), the event funnel (where they drop off), time on page and exits. The hole that shows up first gets fixed first; the rest wait in line.
- Key events: call, form, WhatsApp and primary button click.
- Consolidation: which pages receive traffic and produce no action at all.
- Exits: which page sends the most people away? That is where work starts.
- Typical mistake: optimising the pretty middle page instead of the leaky front door.
80% of the gains come from the deep work
After hundreds of optimisations, what pays most is not micro-buttons: it is the deep mistakes. An unclear pitch, a hidden price, trust without visible address or phone, an eight-field form. Good CRO first removes the most painful blocker and only then plays with colors.
- Clear pitch in 5 seconds: what you sell, to whom and what to do.
- Visible trust: address, phone, real cases and objections answered.
- Zero friction: short forms, fast load, readable on mobile.
- One goal per screen, with everything pointing there.
How to prioritise without arguing
When you have ten things to improve, sort them with two questions: what is their impact on the goal and how cheap is it to test. The matrix is simple: cheap and high impact gets done now; expensive and no impact gets discarded; expensive and high impact splits into small tests before committing.
The four moments of truth
Every website converts in four moments: the first screen (the pitch), trust (who you are and why to believe you), the action (the button and the form) and the exit (what happens if they do not convert). Attack all four and the funnel will be healthy; each one is a line of work with its own measurement.
- First screen: won or lost in five seconds.
- Trust: address, phone, cases and answered questions.
- Action: one clear primary button and a short form.
- Exit: a rescue lead — email, WhatsApp or content that keeps the undecided from leaving.
Why results take real time to show
CRO is not a switch: it is a process. Every change needs to be measured with enough traffic for the result to mean something, and that takes weeks, not days. A large fix can show a signal in two or three weeks; a small one can take longer. Whoever promises to double conversions in a week is telling a story. Whoever shows you the plan, the hypothesis and the decision metrics is doing the real work.
At the Lab we document every experiment: hypothesis, what we did, what happened. If your website has traffic and does not convert, CRO is the highest-return lever you have left to pull.
The proof is the text. Published from the Lab.