What we found auditing 228 local business websites

Over the past months we audited 228 local businesses across Romania, Germany, Austria, Italy, Spain and Poland, in eighteen cities. Restaurants mostly — 146 of them — plus 41 dental practices, 33 dermatology clinics, seven ophthalmology clinics and one law firm. Every one was scored on the same rubric we use for the free audit: does the site work on a phone, can a search engine read it, can a customer book or order, do the basics agree with each other.

Here is what the pile looks like when you stack it up.

Every single one already had a website

Not one of the 228 was missing a website entirely. This is the finding that reframes everything else, because the common assumption — that local businesses are offline and need to get online — is a decade out of date. They are online. They have been online for years. Someone built them something.

The average digital presence score across all 228 was 25.8 out of 100. Two hundred and six of them, just over 90 percent, scored below 40.

So the gap is not presence. The gap is that what exists does not work.

They are well liked and badly represented

The average Google rating across the businesses we audited was 4.4 stars. These are not struggling businesses with an unhappy customer base. They are, by the only public measure available, good at the thing they do.

Thirty-eight of them held ratings of 4.5 or better while scoring below 40 on digital presence. That is the shape of the whole problem in one line: a business people love, that a stranger cannot find, cannot read on a phone, and cannot book with.

Reputation and discoverability turn out to be almost entirely unrelated. Word of mouth built the rating. Nothing built the website.

Nobody can book anything

Of the 228, 197 had no online booking or ordering of any kind. No reservation form, no order flow, nothing but a phone number and, in a good case, an email address.

Broken out by trade, the pattern is sharper than the average suggests:

• Restaurants — 118 of 146 had no online ordering or reservation system. Many were running bookings through Instagram direct messages, which works exactly as long as somebody is holding the phone.
• Dental practices — 40 of 41 had no online booking. One.
• Dermatology clinics — 32 of 33.

The clinics are the striking ones. A dental appointment is a considered, scheduled, high-value transaction that a patient typically decides on in the evening, at home, after putting it off for weeks. There is no worse moment to be told to call between nine and five.

The pattern behind the numbers

Once you have read a couple of hundred of these, the failures stop being varied. Four things account for most of the score gap.

The content is in a format nothing can read. Restaurant menus as PDFs or, worse, as JPEGs of a printed card. Clinic treatment lists as a single page of thirty items. A search engine cannot extract a dish and a price from an image, and neither can an assistant when somebody asks it where to eat nearby. The information is on the site and is simultaneously not on the internet.

The site is in the wrong language. We found clinics running English-only websites for patients who search exclusively in their own language. This is not a small penalty. It is the difference between competing and not appearing.

The facts disagree with each other. Opening hours on the site, on Google and on the door, all different. An address that does not match the map pin. A phone number that changed two years ago and survived in three places. Each of these is individually trivial and collectively fatal, because the one thing a local search has to do is resolve a business to a place and a time.

Nothing is structured. Almost none of the 228 carried usable structured data. No opening hours markup, no menu markup, no medical or local business schema. This used to be an optimisation. It is now the difference between being quotable by an AI assistant and being invisible to one.

What this means if you run one of these businesses

The useful conclusion is not that you need a new website. Sometimes you do. Often you need three specific fixes to the one you already own, and we have told people that and lost the sale.

What is worth checking today, in about ten minutes:

1. Open your own site on your phone, on mobile data, not on the café wifi. Time it. Try to find the menu or the price list. Try to book.
2. Search your business name plus your city in an incognito window. See what Google shows above your site, and whether the hours are right.
3. Ask ChatGPT or Perplexity for the best place in your category in your neighbourhood. See whether you are mentioned at all.

The third one is new, and it is the one that surprises owners most. A business can rank respectably in Google and be completely absent from the answer an assistant gives, because the assistant needs to be able to read and cite specific facts — a dish, a treatment, an opening time — and a PDF menu offers it nothing to work with.

Method, and its limits

These are the 228 businesses we audited, not a random sample of European local business. They were selected because they looked like prospects, which biases the set toward businesses with visible room to improve. The digital presence score is our own rubric, not an industry standard. Google ratings were taken at audit time and move.

What the set is good for is the pattern, and the pattern is consistent enough across six countries and five trades that we would be surprised to see it break: the businesses are good, the reviews are good, and the thing standing between them and a stranger with a phone is a website nobody has looked at properly in years.

If you want yours looked at properly, the audit is free and takes 48 hours.