A website rarely loses people because it looks wrong. It loses them at the moment the next step stops being obvious — when someone has understood roughly what you do, but cannot tell what they are supposed to do about it.
That moment is usually earlier in the page than anyone expects, and it is nearly always fixable without a redesign. What follows is the order we work through when a site is getting traffic but not enquiries.
The key idea
Most conversion problems are navigation problems in disguise. Before you change how a page looks, find the exact screen where the next step stops being obvious.
Start with what the page is for
Every page should have one job, and you should be able to write it as a single sentence: this page exists so that a particular person can do a particular thing. If you cannot finish that sentence, the page is the problem, not the traffic.
The sentence does real work because it is specific. “This page exists so a facilities manager can see whether we handle multi-site contracts and ask for a quote” produces a different page from “this page exists so visitors can understand our range of services.” The first has an obvious ending. The second does not have an ending at all.
Write the sentence at the top of the draft and remove anything that does not serve it. Supporting detail is not deleted — it moves below the decision rather than in front of it.
Where people actually drop out
Three patterns account for most of it.
- A hero that describes the company rather than the reader’s problem. “Digital solutions for ambitious businesses” could belong to thousands of companies, so nobody reading it recognises themselves.
- Navigation that offers ten equal choices where one obvious one was needed. Equal weight is not neutrality — it hands a decision to somebody who has had four seconds of context.
- A form that asks for more than the visitor is ready to give. Asking for a phone number at first contact is a materially bigger request than asking for an email address.
The second-click problem
Often the first click works and the second one does not. Someone clicks “Web design” and lands on a page that starts again from company history rather than picking up where the click left off. Each page should continue the sentence the previous one started: if the link promised something, the first screen of the destination should deliver on it before it does anything else.
How to find the moment on your own site
You cannot see this on a page you designed, because you already know where everything is. Three things that do work:
Take the page away. Show the top of it to someone outside your business for five seconds, then hide it and ask what the company does and who it is for. That five-second test is the spine of what your homepage needs to say in the first five seconds, which goes into it properly.
Watch a phone, not a desktop. The first screen on a phone is far smaller than the one the page was designed on, and it is where most people will meet you.
Look at engagement per page, not site totals. In Google Analytics 4, engagement rate is the percentage of engaged sessions, and a session counts as engaged if it lasts longer than ten seconds, includes a key event, or has at least two page or screen views. Bounce rate is the inverse. Neither number is a verdict on a page — one that answers a question quickly and sends someone away satisfied will look poor by both — but the pages sitting well below your own site average are the ones to look at first. That is what analytics and reporting is for: finding the screen that is losing people, rather than producing a dashboard.
If a visitor cannot tell what to do next within a few seconds, the design is doing the wrong job.
Make the next step physically easy
Once the next step is obvious, it still has to be easy to hit. WCAG 2.2 sets a minimum target size of 24 by 24 CSS pixels at Level AA, with an exception where targets are spaced far enough apart that a 24-pixel circle centred on each one does not overlap its neighbours.
Treat that as a floor rather than a goal. In practice: one primary action per screen, styled so it is visibly primary; secondary options that look secondary; enough space around a tappable thing that a thumb cannot catch the wrong one; and on a long page, the primary action repeated once further down rather than existing only at the top.
Speed and stability are part of the path
A slow page breaks the path as effectively as a confusing one. Google’s Core Web Vitals give three measurable checks: Largest Contentful Paint at 2.5 seconds or less, Interaction to Next Paint at 200 milliseconds or less, and Cumulative Layout Shift at 0.1 or less, each assessed at the 75th percentile of page loads.
Cumulative Layout Shift is the one that matters most here, and the one most often ignored. A layout that jumps while it loads moves the button out from under a thumb that was already on its way to it. That is a path problem wearing a performance costume, and it is usually a web development fix rather than a design one.
What to change first
You do not need a redesign to start. In order:
Where to start
- Write the one-sentence job for the page.
- Cut or demote everything on the first screen that competes with it.
- Make the primary action unmistakable, and big enough to hit comfortably.
- Open the first screen on a phone and check it still does all three.
- Come back a fortnight later and judge it on engagement, not opinion.
This is the thinking behind how we approach web design: structure first, then interface, then the visual layer. A beautiful page built on an unclear structure is still an unclear page.
Weighing up a rebuild? It is worth checking whether the structure is the real problem first.
