Most conversion advice starts with the button. Make it orange. Move it up. Change the wording to something more urgent.
Sometimes that works, and when it does, nobody can explain why, which means it cannot be repeated. The reason is that the button was never the problem. The problem was that somebody arrived wanting to know whether you handle their situation, could not tell, and left before reaching any button at all.
User-centred conversion optimisation starts one step earlier than the tactics: with what the person came to do.
User-Centred Conversion Optimisation: Clearer Journeys
Find out what they actually came for
Not a persona with a stock photograph. The specific thing somebody was trying to achieve when they landed on your page.
Three sources answer this cheaply. Your support inbox and sales calls, which record the questions people ask repeatedly. Your site search queries, which record what people could not find. And watching five people attempt a real task while you say nothing.
That last one is the most informative and the least done. Five sessions will usually surface the same three problems, which is enough to act on.
Then find where the journey breaks
Count the pages somebody loads between arriving and completing your primary action. If it is four and could be two, every additional step is a place to lose them.
Look at drop-off by step in your analytics. Analytics tells you where people left; observation tells you why. Neither works alone.
Common breakages are unglamorous: information people had to hunt for, a form asking more than the task requires, a page that answers slowly, navigation labelled around your internal structure rather than their vocabulary.
Fix the friction before the persuasion
Put the frequently wanted facts where they need no searching. What you do, who it is for, what it costs where you can say, and what happens after contact.
Reduce the form to what routes the enquiry. Every field is answered by everybody who contacts you, so a question useful once in twenty is paid for by the other nineteen.
Validate inline rather than on submission, so nobody retypes anything after an error.
One primary action per page, placed where the decision gets made rather than only at the bottom. Somebody convinced by the third paragraph should not have to scroll past four more.
Accessibility is conversion work
An enquiry form that fails keyboard navigation turns away somebody who had already decided to contact you. That is a lost sale, not a compliance matter.
Work towards WCAG 2.1 Level AA: sufficient contrast for reading outdoors, visible focus states, descriptive links, and error messages explaining what to fix.
Test the enquiry journey specifically, with a keyboard and at 200 per cent zoom. That is the journey with revenue attached.
Speed belongs in the same conversation
A page nobody waits for converts at zero regardless of how well it is written.
Test on a real mid-range phone on mobile data. Where performance and environmental impact both apply, treat them as the same lever: reducing unnecessary data transfer and processing supports a faster experience while lowering estimated digital carbon, though environmental figures remain modelled rather than directly measured.
Test against outcomes, not opinions
Change one thing at a time and measure enquiries rather than clicks. A button producing more clicks and fewer qualified enquiries has made things worse.
Give changes long enough to produce a meaningful result. Small businesses rarely have the traffic for statistical significance in a fortnight, and pretending otherwise leads to confident decisions built on noise.
Where volumes are low, qualitative evidence is more honest than a split test. Five people struggling with the same field tells you more than a two per cent difference across ninety visits.
Our sustainable web design work treats research, accessibility and speed as one brief rather than three projects.