Ask how a small business’s website is doing and you usually get a score. Clikiti gives one too: usability, speed, accessibility, search and trust, out of 100. Across 3,210 websites in 20 industries, from dentists to wedding venues, the scores barely move. Every industry’s median lands between 74 and 80.
A score like that says the site works. It does not say whether the site helps anyone decide. So each report also asks the questions a customer brings: six to ten per industry, written for that industry, each one passing only when the scan finds the answer on the home page or the pages it links to. The questions are reported on their own and do not feed the score.
The questionWhat will it cost?
Nineteen of the twenty reports ask some version of it, in the industry’s own words. In the typical industry, 29% of sites answer. Schools and tutors do best, at 68%. Optometrists do worst: 7% show the price of an eye exam, a contact lens fitting or a pair of glasses.
Who says what it costs?
Share of sites that answer each industry’s price question, in that industry’s own words. The typical industry is highlighted. The insurance report asked no price question.
A score of 77 says the site works. It doesn’t say what anything costs.
The silenceEvery industry has its own.
Price is the common gap, but each report found the one question its industry leaves most often unanswered. For salons it is access: 3% say whether a client in a wheelchair can get. For wedding venues it is the calendar: 9% show which dates are still free. For vets and dentists it is whether they are taking new patients at all.
The question each industry skips
The least-answered question in each of the 20 reports, and the share of sites that answer it.
- 3%Salons“Can I get in with a wheelchair?”
- 7%Optometrists“What will an eye exam cost me?”
- 9%Schools and tutors“Can I reach a person?”
- 9%Wedding venues“Is our date still free?”
- 12%Pest control“Who will treat my home?”
- 13%Accountants“What will it cost me?”
- 13%Funeral homes“Can you honour our faith or culture?”
- 13%Real estate“What will it cost me?”
- 15%Landscapers“Roughly what will it cost?”
- 17%Home trades“What if it fails again?”
- 17%Vets“Are you taking new patients?”
- 18%Physio and chiro“What if I pay myself?”
- 19%Movers“Is the price fixed?”
- 20%Gyms“Where do I park, and can I bring my kids?”
- 22%Dentists“Are you taking new patients?”
- 22%Insurance agencies“Are you licensed where I live?”
- 29%Auto shops“Roughly what will it cost?”
- 29%Restaurants“Can I eat here with my allergy or diet?”
- 34%Med spas“What will it cost?”
- 39%Law firms“Can I reach someone now?”
The looksTwenty industries, twenty reports.
A report only gets cited if people read it, so none of them share a template. The quality bar is the same for all twenty; the style is the industry’s own. Accountants get a ledger, optometrists an eye chart, funeral homes an order of service, wedding venues an invitation. Each look sets its own type, palette, imagery and motion.








Report spreads from the published previews. The photography is AI-generated; each report’s closing note says so.
Why openThe pitch starts with a number.
The reports are free under CC BY 4.0, with the data behind them. That is the point. An agency can open a call with a number about its prospect’s industry, and an assistant asked how good law firm websites are can answer with a Clikiti figure and a link. Every report says how it was made: the sample, the date, the rules and its margin of error.
The record
What every report ships with.
The auto shops report: 29% show any price, a service, a fee, a labour rate or a car.
- A web page, a PDF and MarkdownTwenty of each; the PDFs alone come to 82.8 MB.Clikiti Research index
- The dataEvery share in the report, as JSON and CSV.data.json and data.csv per report
- A citationReady to paste, with the licence: CC BY 4.0, credit Clikiti with a link.Each report
- Its margin of errorFrom ±7.0 to ±9.2 points at 95% confidence, read as estimates for the whole industry.Each report’s method
- No site singled outResults are aggregates; only the best-in-class example is named with its results.Each report’s method