Example: How a Local Clinic's Site Could Go from Invisible to Found

TL;DR: This is a worked example with a made-up business, Harbor Dental Clinic, and illustrative numbers. Its old site scores 41 out of 100 on our check: one long page for every treatment, opening hours that disagree with Google, a heavy photo slider and nothing that says what the clinic does and where. Splitting the treatments into their own pages and fixing the basics takes it to 68. A second round of smaller fixes, mostly labelling content that already exists, takes it to 82. Run the free Website health check to see where your own site starts.

We wrote this example because the question we hear most from business owners is not "what is my score" but "what would you actually fix first, and would it be worth it". A real client's report answers that for one site. A made-up clinic lets us show the typical path without publishing anyone's business, and without pretending the numbers were measured.

Who is the example clinic?

Harbor Dental Clinic is fictional. Picture a family dental practice with two dentists and a hygienist in a mid-sized town, open six days a week, with most new patients coming from word of mouth and from Google Maps. The website was built from a template four years ago by a relative who has since moved on to other hobbies. Nobody has touched it since, apart from changing the phone number once.

The owner's complaint is familiar: people who already know the clinic can find it, but searches like "dentist open Friday" or "teeth whitening near me" show three competitors and never this clinic. The owner also noticed that when a friend asked ChatGPT for a dentist in town, the clinic did not come up at all.

What did the first check find?

The first check scores the site 41 out of 100. Nothing is broken in the way a crashed site is broken. The site loads, the phone number works and the photos are nice. The problems are all in what Google and AI assistants can read:

• One page for everything. Check-ups, fillings, crowns, whitening and children's dentistry all sit on one long home page. A search for "teeth whitening" has no page to land on, so Google picks a competitor that has one.
• Opening hours that disagree. The site says the clinic closes at 18:00 on Thursdays. The Google Business Profile says 19:00. Google notices mismatches like this, and so do patients who turn up at 18:30.
• A heavy photo slider at the top of the page. Five full-size photos load before any text on a phone, so the main content takes about four seconds to appear on a mid-range phone. Google's own guidance on loading speed treats anything over 2.5 seconds as needing work.
• No sentence that says what and where. The main heading is the slogan "Smiles that last". The words "dentist" and the town's name appear only in the footer.
• Nothing labelled for Google. There is no structured data at all, the hidden layer of code that tells search engines this is a dental clinic, at this address, open at these hours.

A patient on a fast laptop sees none of this. A search engine sees all of it, and so does every AI assistant that reads the page to answer "which dentist near me is open on Friday".

Which fixes came first, and why?

The report sorts every finding by how much it holds the site back and how much work it takes. In the example, the first round is the big structural one:

The redirects are the fix owners most often skip, and they matter. Google's guidance on moving pages is plain that permanent redirects are how a new page keeps whatever the old one had earned.

After this round the example site scores 68. Most of the jump comes from content and on-page, which carry 43% of the weight between them and were both scored on a single page that tried to be about everything.

What moved the score from 68 to 82?

The second check, a month later in the example, finds a site in decent shape with a list of smaller, cheaper gaps. This is the usual shape of a second report: the structural problems are gone, and what remains is specific.

• The clinic has 14 questions and answers written on its pages, such as whether it sees children and what happens at a first visit. None is labelled as a question and answer, so it gets labelled.
• 22 patient reviews sit on the home page as plain text. They get labelled, and the clinic starts asking patients to leave the same review on Google.
• The price list is a table of numbers with no labels. Each treatment gets a labelled price range.
• The Google Business Profile has two photos and no treatment list. The clinic adds both, following Google's tips for a better profile.
• The AI assistants that answer local questions were already allowed in, but each treatment page takes three paragraphs to reach the point. The first two sentences of each page are rewritten to answer the obvious question directly.

Each of these is an afternoon of work or less. Together they take the example site to 82.

How do the category scores add up?

These are illustrative numbers, but the arithmetic is the real method. The overall score is the weighted average of the seven categories, as described in our 2026 SEO audit checklist: the first row comes to 41.3, the second to 67.8 and the third to 82.1.

What is still open at 82?

A good score still comes with a to-do list. In the example, the third report leaves these open, in the order we would do them:

1. Prove the booking form works. The online booking form sends requests to another company's booking system. From outside we can see it submit, but we can't see the request arrive. One real test booking settles it, and it is the page that brings in new patients.
2. Move reviews to Google. 22 reviews on the site and four on the Google Business Profile is the wrong way round for local search.
3. Answer questions people ask before they choose. Every page describes the clinic's own services. Nothing answers "does whitening hurt" or "when should a child first see a dentist", which is how someone who has never heard of the clinic finds it.
4. Add the dentists' full names and qualifications to each treatment page. Health topics are held to a higher standard of trust.

The first two are hours of work. The third is the slow, strategic one, and it is the reason a good score is not the same thing as a finished site.

What can a real small business copy from this?

Three things transfer to almost any small site. First, give every service its own page. A one-page site gives Google and AI assistants one chance to match one search, and that is rarely enough. Second, check the boring facts in every place they appear: name, address, phone and opening hours on the site, on Google and anywhere else you are listed. Third, measure before and after with the same method, so a change in the score means a change in the site and not a change in the ruler.

You can start today with the free Website health check on your home page and the Speed test on your phone. If you want the whole site checked, with a score, the problems in order and a fix list you can hand to anyone, that is the full report. You can also see what the report itself looks like on the example report page.