Should You Fix Your Homepage or Your Services Pages?

Almost every owner has a homepage they are quietly unhappy about. The main photo is four years old, the wording was written in a rush the week before launch, and somewhere in the middle there is a sentence about a passion for quality that nobody, including you, has ever finished reading. So the homepage stays at the top of the list. It gets redesigned, re-photographed, and rewritten, and the pages that describe what you actually sell get whatever attention is left over, which is usually none.

This is an expensive habit, and it is easy to break, because you do not need an opinion to settle it. Two ordinary reports in your analytics will tell you where people go and where they give up, and between them they will point at the one page that deserves your next free afternoon. In my experience it is very often not the homepage.

Think about how people arrive. Someone searches for "emergency plumber Bristol" and Google sends them to your emergency call-out page, not your front door. Someone clicks the link in your Instagram profile and lands on a wedding cakes page. Someone gets your details from a neighbour, types your name into Google, and yes, that person does land on the homepage. Three visitors, three routes, and only one of them ever sees the sentence about your passion for quality.

The homepage has a real job, and it is a narrow one: tell a stranger who you are, roughly what you do, and where to go next. That is it. The persuading, the pricing, the reassuring, the answering of the awkward question about lead times — that all happens on the pages further in. When owners pour their effort into the homepage, they are usually polishing the signpost while the shop floor stays exactly as it was in 2021.

Your top pages report is simply a list of the pages on your site ranked by how many times they were viewed. It is the least glamorous report in analytics and the most useful. Open it, set the date range to the last ninety days so you are not reading noise, and look at the first eight or ten rows.

Two things are worth noticing. First, what proportion of all views the homepage takes. Across small business sites I would loosely put it somewhere between a fifth and a half, and the busier and better-established your search presence, the lower that share tends to be. Second, and more interesting, which non-homepage page is highest. That page is doing real work for you, probably without ever having been edited since it was written, and it is your strongest candidate for attention.

If your top pages list is dominated by one service, that is a signal about what the market wants from you, not just what your site says. A solicitor who assumed probate was a sideline and finds the probate page pulling three times the views of the conveyancing page has learned something worth more than a new hero image. The top pages report is where that lands first.

An exit page is the last page someone looked at before they left your site. Every visit has exactly one, so exits are not automatically bad — somebody has to leave somewhere, and if they leave from your thank-you page after booking, that is a success. What matters is the pattern: which pages leak people who had clearly not finished what they came to do.

This is different from a bounce, which is when someone arrives, looks at one page, and leaves without going anywhere else. A high exit rate on a page deep in your site means people got that far, read it, and then stopped. That is the more painful number, because those visitors had already given you their interest. Something on that page took it back. The exit pages report is where you find them.

Read the two reports side by side. High views and low exits means a page that is doing its job. Low views and high exits usually means a page nobody needed. High views and high exits is your problem page, and it is worth more of your time than anything else on the site.

Here is the shape of a report I would call unremarkable — the numbers are invented for illustration, but the pattern is one you will recognise. Ninety days of data for a small building firm:

The contact page's 71% exit rate looks alarming and is almost certainly fine — people find the phone number and leave to ring it. That is the whole point of the page, and it is a good reminder that no single number means anything without knowing what the page is for.

The kitchen extensions page is the real story. It is nearly as busy as the homepage, and six in ten visits end there. Those are people who searched for exactly what you sell, arrived at the page describing it, read it, and left. They did not go to contact. They did not look at your gallery. Whatever was on that page either answered their question badly or answered it so completely that they had no reason to go further — and you can usually tell which by adding a scroll depth measurement, which records how far down a page people actually get before they stop. If most readers stop a third of the way down, the page is losing them mid-sentence. If they reach the bottom and leave anyway, the ending is the problem: no phone number, no next step, no sense of what happens if they get in touch. Scroll depth by page turns that from a guess into something you can see.

Meanwhile the homepage is quietly fine. It sends people onward. Repainting it would change nothing about the 1,760 people arriving at the extensions page with money in mind.

A page can leak visitors for reasons that have nothing to do with the words on it. It might take eight seconds to load on a phone because of an uncompressed photo of a finished kitchen. It might have a broken link to a gallery that no longer exists. The enquiry form might be failing silently on mobile. I have seen all three, and in each case the owner's instinct was to rewrite the copy, which fixed nothing.

So check the plumbing first. A quick technical read of the site will tell you whether your busiest page is slow, broken, or missing from Google's index, and that takes minutes rather than an afternoon. If you have not done that recently, our free report card will give you an outside read on speed, links and basic search visibility without you having to install anything.

Once the mechanics are sound and the page still leaks, then it is a content problem, and then your afternoon is well spent on it.

Analytics can show you that someone left your kitchen extensions page. It cannot tell you they left because your work looked too expensive, or because they saw the word "extensions" and realised they meant "renovation", or because their toddler woke up. Intention is invisible. Anyone who tells you their tool reads minds is selling something.

It also cannot see what happens off the site. If a visitor reads your services page, writes your number on the back of an envelope and rings you two days later from a different phone, that shows in your analytics as a person who left and never came back. It shows in your bank account as a job. This is why the exit report is a prompt to go and look at a page with fresh eyes, not a verdict on it.

The one thing you can close some of that gap with is a defined goal — a form submission, a click on the phone number, a booking confirmation — so that at least the actions taken on the site are counted rather than inferred. Setting up conversion tracking on the two or three actions that matter is a half-hour job and it changes every other report you read afterwards.

Pull ninety days of top pages and exit pages. Identify the busiest page that is not the homepage, and the page with the highest exit rate among those that get real traffic — ignore anything with a handful of views, because a 100% exit rate on four visits tells you nothing. If those two are the same page, you have your answer and you can stop reading.

Fix the mechanics on that page, then fix the ending. Most service pages fail at the bottom, not the top: they describe the work well and then simply stop, with no clear instruction about what to do next and no reassurance about what happens when you get in touch. Adding a specific next step — a phone number, a short form, a line about how quickly you reply — is a smaller job than a redesign and usually moves the number more.

Then leave the homepage alone until the services pages have had a turn. Come back in a month and read the same two reports. If the exit rate on that page has fallen and enquiries have not, you have learned something useful too, and you can move to the next page down the list.

If you would rather have this pointed out to you than go looking for it, that is roughly the job our built-in AI advisor does: it reads your top pages, exit pages and goals each week and sends a short ranked list of what to change first, with the numbers behind each suggestion. You can book a short demo and see it running against a real site rather than a screenshot.