What does “working” actually mean?
It means enquiries, and nothing else. A website for a local service business has one job: turning someone who was looking for help into someone who contacts you. Calls, form fills, emails. Visits are not the job. A pretty design is not the job. Compliments from your brother-in-law are definitely not the job.
That changes what you check. A site can look ten years old and work brilliantly because it says plainly what you do and where, loads fast on a phone, and has a number you can tap. A site can look beautiful and produce nothing because it says “quality craftsmanship, trusted service” and never names a town. Judge yours by the phone ringing, not the paint job.
It also means the anxiety is answerable. “My website does nothing” is a feeling. “Nobody is arriving” and “people arrive and leave” are two different problems with two different fixes, and ten minutes of checking tells you which one you have.
The ten-minute check: seven things you can test tonight
Seven checks, no tools, your phone will do. Do them in order and write down which you fail: that list is your to-do list, and the table below turns each failure into a fix.
1. Search your own business name
Type your business name into Google. Does your website come up first? Is there a business panel with your details, opening hours, reviews and photos, and is everything in it right?
This is the friendliest search you will ever appear in: the person already knows your name, usually from a recommendation. If even this looks wrong — old phone number, no reviews, a competitor above you — you are losing your warmest leads before they reach the website. The fix starts with claiming and completing your Google Business Profile.
2. Check you are in Google’s index at all
Search site:yourdomain.co.uk, with your own domain. That lists the pages Google currently holds. If nothing comes back, Google does not have your site and every other question is academic. If three pages come back and you have twenty, most of your site is invisible. This is the single most common cause of complete silence, and it is covered in full in why isn’t my website showing on Google.
3. Search what you do plus your town, in a private window
Open a private or incognito window, so Google is not flattering you based on your own browsing history, and search the way a customer would: “electrician Wrexham”, “wedding photographer Yeovil”. Do it for each of your two or three main services.
Look for two things: the map block at the top, and the ordinary blue links below it. Missing from the map block is a Business Profile problem. Missing from the blue links is usually a service page problem.
A correction to an earlier version of this page. This guide used to say the map block takes “around 42 to 44% of the clicks on local searches”. That figure is repeated everywhere, but every trail we followed led back to a single small click study rather than to data we would stand behind, so we have removed it. What is safe to say is that the map block sits above everything else on the page, so being in it matters. How much, precisely, nobody has published credibly.
4. Ask an AI tool about your trade in your town
Open ChatGPT, Copilot or Gemini and ask it what a customer would ask: “recommend a good roofer in Barnsley”. Do you exist to it? Do your competitors?
This is no longer a novelty. Ofcom’s Online Nation 2025 report, published on 10 December 2025, records ChatGPT at 252 million UK web visits in August 2025, up 156% on the same month the year before, and finds 53% of people saying they often see AI summaries in search results. Ofcom also reports that around 30% of keyword searches now return an AI-supported overview. If you want the practical version of this check, read how to get your business on ChatGPT.
A second correction. An earlier version said Ofcom found 54% of UK adults use AI tools, and 79% of 16 to 24 year olds. Re-checking Online Nation 2025 in September 2026, we could not find those figures in it, so they have been replaced above with figures we could locate in the report itself.
5. Test your own site on your own phone
On your phone, not your computer, because that is where your customers are. Load the site cold on mobile data, not the office wifi. Is it quick? Can you read a service page without pinching? Tap your own phone number: does it dial? Send yourself a message through your own contact form: does it arrive?
Do that last one properly. Form failures are silent and common, and every one is an enquiry that tried to happen and could not. Check the spam folder of whichever inbox the form sends to while you are there.
6. Run one free speed test
Put your homepage and one service page through Google’s PageSpeed Insights. A minute each, and it gives you two things: field data from real visitors over the previous 28 days, and a lab score from a simulated mid-range phone. Google states that a lab score of 90 or above is good, 50 to 89 needs improvement, below 50 is poor.
7. Ask what is measuring your enquiries
The last one is usually the killer. What, right now, counts how many enquiries your website produces? For most small business sites the honest answer is nothing — which means every opinion about that website, yours included, is a guess dressed up as a judgement.
The fix costs nothing but an evening: Google Analytics to count the people arriving, Google Search Console to see what they searched to find you, and conversion tracking so the calls and form fills are counted rather than estimated.
Symptom, cause and the guide that fixes it
This is the useful part. Find the row matching what you saw and follow the guide in the right-hand column. Most people fail two or three rows, not one.
| What you saw | What it usually means | Where to go next |
|---|---|---|
| site:yourdomain.co.uk returns nothing | Google does not have your site at all — often a blocked setting left on after a rebuild, or nobody ever submitted it | Why isn’t my website showing on Google |
| Your own business name brings up the wrong details, or no business panel | Your Google Business Profile is unclaimed, incomplete, or duplicated | Set up your Google Business Profile |
| You are missing from the map block for service plus town | Profile not verified, category wrong, or no reviews | Set up your Google Business Profile |
| You are missing from the ordinary results for service plus town | One vague “Services” page instead of a page per service | Write a service page that gets found |
| You show up for your town but not the surrounding ones | No location coverage in the wording or the page structure | Create location pages that get found |
| AI tools name your competitors and not you | Your pages never state plainly what you do and where, so there is nothing to quote | Get your business on ChatGPT |
| Nothing in Bing, and Copilot has never heard of you | Bing has never been told the site exists | Set up Bing Webmaster Tools |
| Changes you publish take weeks to appear | Search engines are not being told when pages change | Enable IndexNow for Bing |
| Your listing in the results reads badly or shows the wrong text | Missing or duplicated page titles and descriptions | Write a good page title and meta description |
| The site is slow or awkward on your phone | Heavy images, a bloated theme, or a design never tested on mobile | The Core Web Vitals section below, then whoever maintains your site |
| Your test form never arrived | A broken form — the most expensive fault on this list | Why isn’t my website getting enquiries |
| Traffic arrives but nobody contacts you | A conversion problem, not a visibility problem | Why isn’t my website getting enquiries |
| Nothing is counting anything | No analytics, or analytics with no conversion tracking | Set up conversion tracking |
| The site looks fine and you still cannot see why it is quiet | Usually the words on the pages rather than the build | I paid for a website and get nothing |
What is a good conversion rate for a small business website?
The best UK figure we can point at is 5.13%: Ruler Analytics’ 2026 benchmark, published on 26 May 2026 and built from over 110 million sessions and more than 5 million tracked conversions across 13 industries. Within that, construction and engineering runs at 4.9%, professional services at 6.1%, health and social care at 2.3%, beauty at 3.5%, and retail and ecommerce at 2.4%. For professional services, Ruler splits the conversions almost evenly between forms at 47.4% and phone calls at 52.6%, which is worth knowing before you conclude that your form is broken.
Two caveats. Ruler measures across paid and organic channels, not just a quiet brochure site, so treat it as a ceiling to aim at rather than par. And a conversion rate is only as good as the thing counting conversions: if calls are not tracked, a trade that takes most of its work by phone looks terrible on paper while doing fine in reality.
About the 2 to 3% figure. An earlier version of this page said “around 2 to 3% of visitors typically get in touch”. It is repeated all over the web and may well be roughly right for a small local site, but we could not trace it to a source worth citing, so it has been replaced with the Ruler numbers above. If your own rate sits at 2%, that is not failure; it means you have a number, which is more than most businesses have.
How many visitors should my website get?
There is no honest benchmark, and anyone offering one for a local service business is inventing it. Visitor numbers depend on how many people search for what you do near where you do it, and that varies by a factor of hundreds between a Birmingham electrician and a thatcher covering three villages in mid Wales.
The number that answers the question is impressions in Google Search Console: how often you appeared, clicked or not. The default Performance view shows the last three months. Impressions in the low hundreds means you are not appearing, and no redesign fixes that. Impressions in the thousands with a trickle of clicks means your titles and descriptions. Healthy clicks and a silent phone means the website. Same tool, three different diagnoses.
What is a good bounce rate?
Google publishes no target, and the metric is less useful than its reputation suggests. In Google Analytics 4 the definition is mechanical: bounce rate is the percentage of sessions that were not engaged, and a session counts as engaged if it lasts longer than 10 seconds, includes two or more page views, or triggers a key event.
The problem is then obvious. Someone who lands on your contact page, reads your number and rings you in eight seconds is recorded as a bounce. Someone who wanders three pages and leaves is not. And because a key event counts, your bounce rate changes the day you configure conversion tracking, with nothing on the site having changed. Use it as a smell test on one page, never as a verdict on the site.
How fast should my website load on a phone?
Google’s published Core Web Vitals thresholds, checked on 15 September 2026, are the only figures worth quoting here, and they are specific:
| Metric | What it measures | Google’s “good” threshold |
|---|---|---|
| Largest Contentful Paint (LCP) | How long until the main content appears | Within 2.5 seconds |
| Interaction to Next Paint (INP) | How quickly the page responds when tapped | 200 milliseconds or less |
| Cumulative Layout Shift (CLS) | How much the page jumps about while loading | 0.1 or less |
Two details people miss. Google measures these at the 75th percentile of page loads, segmented across mobile and desktop, so a page passes only when three quarters of real visits hit the target, and your own fast broadband test proves nothing. And the field data comes from a rolling 28 day window of real Chrome users; a small site may have too few samples to report at all, leaving you only the lab score.
If you fail LCP on a phone, the cause is nearly always photographs uploaded straight off a camera at several megabytes each. Cheap to fix, and worth doing before anyone mentions a rebuild.
Which free tools tell you what
Five free tools cover almost everything a small business needs, and each answers a different question. Using the wrong one is how people conclude their site is fine when it is not.
| Tool | The question it answers | What it cannot tell you |
|---|---|---|
| Google Search Console | Did Google index my pages, what people searched, how often I appeared and was clicked | What happened after the click, or why a competitor outranks you |
| Google Analytics | How many people arrived, from where, on which pages, and what they did | Whether an enquiry was any good, or whether it turned into work |
| Bing Webmaster Tools | The same for Bing, plus an AI Performance report launched 10 February 2026 showing citations in Microsoft Copilot and Bing’s AI answers | Anything about Google, or AI tools outside Microsoft’s surfaces |
| PageSpeed Insights | Real-visitor loading data over 28 days plus a simulated mobile score | Whether slowness is actually costing you enquiries on your site |
| Rich Results Test | Which rich results your structured data could generate | Whether they will appear — Google states plainly that it does not guarantee it |
That last row catches people who have paid for schema markup. Google’s own documentation says: “Google does not guarantee that your structured data will show up in search results, even if your page is marked up correctly according to the Rich Results Test.” Marking a page up enables a feature; it does not buy one.
What none of these tools can tell you
They tell you symptoms, not causes. Search Console can show that you are not appearing for “boiler repair Chester”. It cannot tell you whether that is because you have no page for boiler repairs, because the page exists but never says Chester, because competitors have 200 reviews and you have four, or because your platform hides the page behind a script. Analytics can show people leaving your quotes page in five seconds; it cannot say whether that is the price, the photograph, or a form asking eleven questions.
Nor can they tell you what to do first, which is what most people need most. And searches you run yourself are one postcode on one evening: useful, not gospel, because results are personalised and localised.
What to do first, in order
If you failed several checks, the order matters more than the effort. Work down this list.
- Fix a broken contact form or untappable number today. Everything else is pointless while enquiries cannot physically reach you.
- Claim and complete the Google Business Profile. Free, the fastest visible change for a local business, and the thing standing between you and the map block.
- Set up Search Console and Analytics. An evening’s work that ends the guessing, and lets you tell in three months whether anything helped.
- Give each service its own page, with your towns named in ordinary sentences. The biggest lever on the ordinary results, and the same fix that lets AI tools describe you.
- Compress your images. Cheap, quick, and usually the whole of a failing mobile speed score.
- Then, and only then, ask whether the platform is holding you back. Be sceptical of anyone who reaches that conclusion first.
Some of this takes months to show up, and that is normal rather than a sign it failed. Judge search work on impressions over a quarter, not rankings over a fortnight. For the longer view, see whether SEO is worth it for a small business.
Common questions
- How do I check if my website is on Google?
- Search site:yourdomain.co.uk in Google, using your own domain. That lists the pages of yours Google currently holds in its index. Nothing back means Google does not have you and nothing else matters yet. The reliable version of the same check is Google Search Console, which is free and shows which pages are indexed, which are not, and the reason given for each. Do the quick search first, then set up Search Console so you never have to guess again.
- What is a good conversion rate for a small business website?
- Ruler Analytics' 2026 benchmark, built from over 110 million sessions and 5 million tracked conversions, puts the average across 13 industries at 5.13%. Construction and engineering sit at 4.9% and professional services at 6.1%. The widely repeated 2 to 3% figure is not from any source we could stand behind, so treat it as folklore. Use these as a sanity check, not a target, and compare yourself against your own trend.
- What is a good bounce rate for a website?
- Google publishes no target. In Google Analytics 4, bounce rate is simply the percentage of sessions that were not engaged, and a session counts as engaged if it lasts longer than 10 seconds, has two or more page views, or triggers a key event. That means your bounce rate moves when you change what you count as a key event, so it is a poor measure of whether a site is working. Count enquiries instead.
- How many visitors should my website get?
- There is no honest benchmark for this, and anyone quoting one for a local service business is guessing. Visitor numbers depend on how many people search for what you do near you, which varies enormously between a Birmingham electrician and a thatcher in mid Wales. The useful number is impressions in Google Search Console: how often you appeared. If impressions are tiny, you have a visibility problem rather than a website problem.
- How do I check my website traffic for free?
- Google Analytics counts the people who reach your site, and Google Search Console shows how often you appeared in search, what people typed, and how many clicked. Both are free from Google and take under an hour to set up. Bing Webmaster Tools adds the same picture for Bing. Third party traffic estimators guess from sampled data and are unreliable at small business scale, so use your own tools rather than theirs.
- How do I check if my website is mobile friendly?
- Open it on your own phone and try to use it the way a customer would: read a service page without pinching, tap your phone number to see whether it dials, and send yourself a message through your own contact form. Google retired its standalone Mobile-Friendly Test in December 2023, so the automated stand-in is now PageSpeed Insights, which reports mobile results. Your own thumb still finds more faults than any tool.
- How do I know if my SEO is working?
- Look at impressions and clicks in Google Search Console over months rather than days, and check whether the queries listed are ones a paying customer would type. Rising impressions for your service plus your town is progress even before the clicks arrive. Rankings you check yourself are unreliable because results are personalised and localised. If impressions are flat for three months, nothing is working yet.
- Why is my website not getting any enquiries?
- There are only two possibilities and the check tells you which. Either nobody is arriving, which is a visibility problem fixed by a Google Business Profile, service pages and named locations, or people arrive and leave, which is a website problem: a broken form, no visible phone number, a slow page, or wording that never says what you do and where. Check the form works before you change anything else.
Related guides
Sources, all checked 15 September 2026. Conversion rates: Ruler Analytics, “Conversion Rate Benchmarks by Industry”, 26 May 2026 — 110m+ sessions, 5m+ conversions, 13 industries; 5.13% overall, construction and engineering 4.9%, professional services 6.1% (47.4% forms, 52.6% calls). Core Web Vitals and the 75th percentile rule: web.dev, “Web Vitals”. Field data, the 28-day window and the 90/50 score bands: Google, “About PageSpeed Insights”. Bounce rate and engaged sessions: Google Analytics Help, GA4 engagement definitions. Search Console and its three-month default view: Google Search Console Help. Structured data: Google Search Central, general structured data guidelines. AI Performance report: Bing Webmaster Blog, 10 February 2026. AI search use: Ofcom, Online Nation 2025, 10 December 2025. Three figures are deliberately absent because we could not establish them: that the map block takes 42 to 44% of local clicks (traceable only to one small click study), that Ofcom found 54% of UK adults and 79% of 16 to 24 year olds use AI tools (not in Online Nation 2025), and the 2 to 3% conversion rate repeated across the web. All three appeared in the earlier version of this page and are flagged above rather than quietly dropped.