Does the booking-path audit count a phone-only med spa as a finding?

Yes, and it says so plainly. The phone_only class records that the walker found no online booking mechanism on the pages it walked. It is 11 of the 38 reviewed findings. The audit asks whether a stranger can book online, so the absence counts. A practice that chose the phone on purpose can fairly disagree with the label, and the article gives the rate both ways.

What is the booking-path finding rate without the phone-only class?

25 of 140, or 17.9%, once phone_only is excluded from the reviewed findings. The headline rate that includes it is 36 of 140, or 25.7%. The denominator is the same in both, the 140 domains where the walker reached a booking surface, so the only thing the exclusion changes is the numerator.

What does a phone-only booking path look like when the front desk is closed?

The same as it looks when the desk is open. The page offers a phone number and no other route to an appointment, and the site cannot tell whether anyone will answer. The audit walked each site once and did not measure how often a call goes unanswered, so this is a description of the path, not a claim about the phones.

When a med spa takes bookings by phone only

The question, and why it keeps being asked

The question arrives as whether a med spa should offer online booking at all, and it is usually asked by somebody who already has a working phone and a front desk that answers it. That practice has been taking bookings the same way for years. Patients call, a person picks up, a consultation is arranged, and the website’s job is to make the phone number easy to find. Nothing about that is broken.

We audited 221 med spa booking paths across North and South Carolina in August 2026, and this is the one finding class where we think the practice may be right and the audit may be wrong. This article is about how to decide, using what the walk actually saw and conceding what it could not.

What the audit counted

The walker takes the booking path the way a stranger with a phone would: from the homepage, through whatever the site offers, to the screen where an appointment time is picked. What it does and refuses to do is described on the methodology page. One of the things it records is an absence. When it walks a site’s pages and finds no online booking mechanism at all, it files a finding called phone_only.

After the operator review, the corpus holds 38 findings across 36 domains. Eleven of those 38 findings are phone_only. That makes it the second-largest class in the audit, behind the 13 phone-verification walls that came from one vendor’s default configuration and ahead of every class that describes something actually failing on screen.

The definition matters, so here it is without adjectives: no online booking mechanism found on the pages walked. It is not “the phone was not answered”. It is not “the patient gave up”. It is a walker reporting that the path it was given ends at a telephone number and nowhere else.

The rate with it and without it

The audit’s headline is 36 of 140, or 25.7%: the reviewed findings over the walked population, which is every domain where the walker reached a booking surface and something happened there. 140 excludes the 19 domains a vendor’s robots file kept us out of and the 62 where no booking surface was reached at all, because nothing was walked in those cases.

Take phone_only out of the numerator and the rate is 25 of 140, or 17.9%, with phone_only excluded from the reviewed findings and nothing else changed. The denominator does not move, because every phone-only finding came from a path the walker reached the end of. The path was simply a phone number.

Both figures are in print already. We are restating them here because this is the article where the gap between them is the whole argument. About a third of the audit’s headline rate, the difference between 36 of 140 at 25.7% and 25 of 140 at 17.9% with phone_only excluded from the reviewed findings, is made of a finding class that a reasonable owner may have chosen on purpose.

The case for the phone

It is worth making this case properly rather than as a concession.

A phone call reaches a person. That person can ask what the patient is hoping for, steer a first-timer toward a consultation instead of a treatment they have picked from a menu, catch a contraindication before a slot is wasted, and quote a price with the context a price needs. Most of what a med spa sells is not a haircut. It is a decision that a trained person is better placed to help with than a calendar widget is, and a practice that routes every new patient through a conversation is making a clinical and commercial choice, not neglecting its website.

The audit’s own caveats say this. phone_only counts a business model as a finding, and the corpus flags it as the one class a critic can fairly reject. We agree with the critic often enough that we publish the rate without it, in the same breath as the rate with it.

What the audit did not measure is whether the phones behind those findings were answered. A headless browser does not call anybody. One visit per site, inside a single forty-one-hour window, on an unattended machine, and the walk stopped at the number. Everything a phone-only practice does well happens after the point where our data ends.

What the walk did see

There is a second lane in the audit that is easy to confuse with this one, and keeping them apart is the difference between a finding and an assumption.

62 domains were unreachable, meaning the walk never reached a booking surface. Of those, 47 are an absence claim by a detector: nothing on the homepage offered a way to book, and no tap-to-call link was found either. Those 47 are not phone-only practices as far as the audit knows. They are pages where the walker could not find a phone to call or a button to press, which may mean an unusual layout, a white-label subdomain, or a number sitting in an image where a detector cannot read it.

The eleven phone_only findings are different. They sit inside the 140, which means the walker reached the end of a path. What it found there was a way to call and no way to book. That is a narrower and more defensible claim than “this practice has no online booking”, and it is the only one the data supports.

Online booking done badly is also a finding

The decision is usually framed as phone versus online, as if online were one thing. The audit says it is not.

The largest finding class in the corpus is a wall, not an absence. Thirteen of the reviewed findings are a phone-verification screen that stands before any appointment time is shown: a form for a name, an email address and a mobile number, then a code sent by SMS, then the times. All thirteen are one vendor’s default hosted booking flow, on twelve practices. Two more findings are login walls. One is an account wall. Four are booking links that resolve to an error page. Each of those practices offers online booking, and each of them has a path that, on the day we walked it, did not show a time before it asked for something.

So the honest comparison is not a phone number against a calendar. It is a phone number against whatever the practice’s booking vendor shows a stranger first. A practice that switches on online booking and gets a default wall has moved from one finding class to another, and the second one is larger. That is the reason the repair exists as a separate service from the website: most of the time the vendor is already there, and the work is making it show a time.

The question underneath the question

Phone-only practices rarely ask about online booking because they want a calendar. They ask because of what the phone does when nobody is on it.

A phone number is a booking path with a person at the other end, and it is only a path during the hours that person is there. The website is open at nine at night and the desk is not. The page does not change; the number is still there and still correct. What changes is that the one route the site offers now depends on a call being returned in the morning, and the site has no way to say so.

We are not going to attach a figure to what that costs, because the audit did not measure it and nobody should trust a number invented for this paragraph. What we can say is what the page shows a visitor at that hour: a number, and nothing else that books.

That is the gap an assistant like Claire is built to sit in, and we will say the same thing about it here that the chatbot page says: it answers questions and takes bookings around the clock only if there is something for it to book into. An assistant in front of a phone-only path can take a message. It cannot show a time, because the practice has not decided to show one. The decision comes first.

How to decide

Three questions, and none of them needs us.

Would a stranger booking the wrong treatment cost more than a stranger not booking at all? If most of what you sell needs a consultation first, the phone is doing real work and a calendar would undo it. If a third of your bookings are repeat patients rebooking a treatment they have had before, those patients are being asked to call for something a screen could do.
Look at your own booking vendor’s first screen before you decide anything. Most practices already pay for software that can take an online booking. Open it as a stranger, on a phone with no remembered login, and count the screens before a time appears. If the answer is a verification code, you would be trading a phone-only finding for a wall, and the wall is the larger class.
Write down what the site offers at nine at night. Not what the practice does in the morning. What is on the page. If the answer is a phone number and a contact form, then the site’s after-hours booking path is a form fill and a callback, and it is worth deciding on purpose whether that is what you want it to be.

If the answer to all three is that the phone is right for the practice, then it is, and nothing in this audit says otherwise. The finding stays in the count because the audit asks a narrower question than you do.

What this does not measure

The sample is not a survey. The 221 domains came off an outbound prospecting queue, found by metro text searches and filtered by a predicate that excludes practices we had already written about. It is 188 of 221 in the Charlotte area, with the rest in Greenville, Charleston, Sumter and Columbia. Whatever it measures, it biases the finding rate down, and it says nothing about practices outside the Carolinas.

A phone-only finding is a description of a page, not of a business. The walker did not call the number, did not visit during opening hours on purpose, and did not check whether a form on the page gets a reply. A practice with a phone that is answered on the second ring by somebody who knows the treatments has a better booking path than most of the calendars in this corpus, and the audit cannot see that.

73 domains were incomplete, a third of the sample, because the walker’s click grammar stalled at a vendor’s first screen. Nothing was proven about those sites in either direction, and any of them may turn out to be phone-only, walled, or clean. They stay in the 140 as the cases where nothing was proven, which is why the 25.7% and, with phone_only excluded from the reviewed findings, the 17.9% are both lower than the rate you would get by throwing them away.

Every figure is pinned to two dates: measured 27 August 2026, reviewed 7 September 2026. Some of these sites will have changed since.

If you would rather we walked yours, we will do that for free, and you keep the written report either way. If the report says the phone is doing its job, that is what it will say.