What almost every booking page gets wrong

Almost every booking page fails in the same place, and it is not where people expect. The photographs, the opening hours, the phone number and the prices are essentially always there. What is missing is everything that required somebody to make a decision: a cancellation window, a focused catalogue, service names you can tell apart. We scored 62 real salon pages against twelve checks and not one passed all twelve.

That split is the whole article. The software fills in the easy half, nobody fills in the other, and the other half is where clients actually get stuck.

Key takeaways

• Not one of 62 pages passed all twelve checks. The best scored 11 and the median scored 9.
• Everything the platform populates automatically sits at or near 62 of 62.
• Everything requiring an editorial decision sits at a third or less.
• The single worst result: 4 of 62 pages mentioned cancellation at all.
• Fixing your page is not a design project. It is roughly five decisions and five sentences.

The twelve checks

We picked things a client either can or cannot find, so each check has an unambiguous answer:

1. A real photograph rather than a stock placeholder.
2. A description written by the business.
3. Opening hours.
4. A phone number.
5. A price on every service in the catalogue.
6. Any mention of cancellation.
7. A star rating.
8. A team section.
9. A focused catalogue, meaning thirty services or fewer.
10. No near-duplicate service names.
11. No description text repeated across services.
12. Service names not bundled with a plus, ampersand or slash.

The first four are what a booking platform sets up for you. The last four are what only you can decide.

[ORIGINAL DATA] Scoring 62 pages against twelve checks

Methodology. On 19 August 2026 we scored the same 62 salon booking pages audited in what an online booking page needs, drawn deterministically from Fresha's public salon sitemap. Each check was evaluated from the page's own JSON-LD or from a term search of the served HTML, so every result is reproducible. Sample size: 62 pages, twelve checks, one platform. Limitation: server-rendered markup only, so anything appearing later inside the JavaScript booking flow is not counted.

And the distribution of total scores:

The median page passed nine. Seventeen pages passed ten or more, and none passed everything.

[UNIQUE INSIGHT] Sort that table and it splits cleanly in two, with almost nothing in the middle. Six checks sit at 50 or above; five sit at 37 or below; one sits at 4. The dividing line is not difficulty or cost - a cancellation sentence is cheaper than a photoshoot. The line is who performs the action. Everything above it is filled in by the software during setup, or generated from data the business enters for its own operational reasons. Everything below it requires the owner to sit down and make an editorial judgement that nobody prompts them to make and no screen marks as incomplete. That is why the failures are so consistent across sixty-two independent businesses in a dozen countries: they are not making the same mistake, they are hitting the same absence of a prompt. Which also means the fix is not effort. Seventeen pages already do ten of twelve. It is a checklist somebody has to be handed once.

The four things nearly everyone does

Photograph, description, hours, phone. Sixty-two out of sixty-two, every time.

This is worth stating plainly because it changes where effort pays. If you are polishing your photographs to stand out, you are competing on the one dimension where everybody has already arrived. The same goes for hours and contact details.

Prices are nearly there too, at 60 of 62, and a star rating at 50. On the ratings we found something worth knowing: of the 50 rated venues, 46 sat at exactly 5.0 and the rest at 4.9, so the number cannot separate you from anyone.

The five things nearly nobody does

A focused catalogue. Only 37 pages listed thirty services or fewer. A long list is a navigation problem, not a range.

Names you can tell apart. Only 21 pages avoided near-duplicate service names. When two lines look the same, the client either guesses or leaves.

Descriptions that differ. Only 21 pages avoided repeating the same description text across services. Copy-pasted text tells a client that the two services are interchangeable.

Unbundled names. Only 26 pages kept service names free of a plus, ampersand or slash. Bundled names hide whether the parts can be booked separately.

A cancellation window. Four pages. That is the number to sit with.

The last one matters beyond politeness. A stated window is what makes a freed slot recoverable - a cancellation the evening before can be offered to somebody else, and a no-show cannot. It also affects attendance directly: across 46,655 appointments, no-shows ran at 9.1% (lead-time analysis, 2015) within a two-week horizon and 38.3% (same analysis) at six months.

The fix, in five decisions

None of this is design work. Each item is one decision and one sentence.

1. Write the cancellation window. "Free to cancel up to 24 hours before; after that the slot is charged."
2. Cut the catalogue to what you want to sell.
3. Take your two most similar services and state the difference in one clause.
4. Split any name containing a plus, ampersand or slash.
5. Say whether each booking confirms instantly or becomes a request.

Then move deposits, lateness and preparation into the confirmation message, where they are read by someone who has already chosen. That channel does real work: across 29 studies, median non-attendance fell from 23.1% (systematic review, 2011) to 12.5% (same review) once reminders existed.

Definitions used in this article

A booking page is the public page where a client selects a service and a time without messaging anyone, such as a profile link or a QR route.

A near-duplicate name is two services named similarly enough that a client must guess between them.

A bundled name is one service name joining two deliverables with a plus, ampersand or slash, which hides whether they can be booked separately.

A cancellation window is the period before an appointment in which a client may cancel without a charge, for example twenty-four hours.

Table stakes are the elements every competitor already shows, which means they no longer differentiate you.

FAQ

What do most booking pages get wrong?

The parts requiring a decision. In our sample, four of sixty-two pages mentioned cancellation, and only about a third avoided duplicate service names or bundled names.

What do most get right?

Photograph, description, opening hours and phone number, at sixty-two of sixty-two. Those are table stakes rather than advantages.

How many pages were actually complete?

None. Two scored eleven of twelve, the median was nine, and the lowest was six.

Is the star rating worth working on?

Not on this evidence. Of fifty rated venues, forty-six sat at exactly 5.0 and four at 4.9.

How long does fixing this take?

An evening. Five decisions and five sentences, which is why seventeen pages already pass ten of the twelve.

Where should deposits and lateness go?

The confirmation message, not the service listing. They are read by someone who has already chosen.

Where this fits

The full seventeen-question checklist is in what an online booking page needs. Writing the text under each service is covered in how to write a service description, and the pricing decision in should you show prices online. The build order for a first version is in set up your booking page in one evening.

The Belom answer to these gaps is online booking.

Editorial note: written by the Belom product team from how the product actually works, not from vendor marketing. More about Belom, or contact us with a correction. Reviewed 19 August 2026.