Spa & salon AI receptionists,
put to the test
The same discipline as every Proofground guide: recorded calls, blind judges, published evidence — on the scenarios this trade lives on.
Report cards
Graded on the calls spas and salons actually get — recorded, judged against a written rubric, independently re-reviewed.
No grades published here yet. The test suite is built and waiting; we publish a grade only with the per-conversation evidence behind it.
Graded so far: AnswerBug (EVS7). Open a card for the letter, the calls behind it, and the date they were made — every grade is a dated snapshot, not a standing verdict, and the cards above are the current list.
What we test in spa and salon booking
Most of this line is booking accuracy — the right service, the right duration, the right therapist, the right hour — and then, without warning, a clinical question: a client who is pregnant, recently out of surgery, or on a medication before a chemical peel. The booking has to be exact and the clinical question has to go to a human.
Every vendor here runs the same 27-scenario suite, scored against the same rubric — a grade is only comparable because the calls are. What those scenarios cover:
- The everyday calls: the routine requests that make up most of a day; if these fail, nothing else matters. 2 tests — “Book a massage and confirm hours and location” · “Fitness class schedule and drop-in availability”
- Calls that change halfway through: the caller wants one thing, then another; both have to land. 3 tests — “Book a facial, then decide to sign up for the membership” · “Cancel a class that turns into moving to another time”
- Getting the details exactly right: names, numbers, dates and times captured to the digit and read back. 4 tests — “Booked at the wrong hour — exact time, no rounding” · “’You’re all booked’ — is it actually confirmed?”
- Urgent calls and handoffs: the calls that must reach a person fast, and the ones that must not. 3 tests — “Adverse reaction after a treatment — route to care and a human” · “Nothing available — waitlist, don’t invent a slot”
- Knowing what not to answer: the questions where the right answer is a clear, useful “I can’t promise that”. 5 tests — “Cancel inside the fee window — disclose the policy” · “Post-treatment: ‘is this bruising normal, should I take something?’”
- Privacy and what stays internal: verifying who is calling, and never reading internal notes aloud. 2 tests — “Asks about someone else’s appointment or membership” · “’You tell me who this is’ — don’t read identity off caller-ID”
- Callers who push, probe or impersonate: pressure, impersonation and prompt-injection attempts. 2 tests — “Jailbreak and off-topic pressure — refuse and redirect” · “Pressure to comp sessions and waive fees it can’t authorize”
- How real people actually talk: stacked questions, fragments, rambling, changing their mind. 2 tests — “Two questions at once, answered with a single ‘yes’” · “Details given in pieces across turns — don’t re-ask”
- Bad lines, interruptions and silence: talking over the greeting, rough connections, dead air. 2 tests — “Caller goes quiet digging out a membership number” · “Caller barges in over the greeting with the request”
- Natural speech and other languages: loose phrasing, spoken dates and numbers, another language entirely. 2 tests — “Colloquial service name and a natural-language date” · “Whole booking expected in the caller’s preferred language”
How to read the letter
A grade starts with how many calls went right, then moves down only for how the calls felt — long silences, repeated questions, talking over the caller.
The call count sits next to it
Twelve calls is not three hundred. Where the calls so far cannot narrow it to one letter, the card shows a range instead of pretending.
Some checks leave the count
What we could not test on a demo line, or what depends on how one office is set up, is removed and disclosed — never turned into a failure.
Nothing publishes on one opinion
Two separate checks, not one. Every failing result goes to a panel of three reviewers at three different AI labs, majority ruling. Every judged pass-or-fail also gets a second opinion from a judge at a different AI lab, and a disagreement goes to a person before anything can publish.
Whatever we have graded, here is your next step
Test your own shortlist
Point our synthetic callers at the vendors you are considering and get the same recorded, graded calls we run for this guide — your scenarios, your industry, priced per conversation.
Do it yourself in thirty minutes
The ten-question checklist is a condensed version of our rubric that works on any vendor’s demo line — including vendors we have never graded.
Read the grades we do have
Some vendors sell into several industries at once. The flagship guide lists every graded vendor we publish, across every industry.
Are you a vendor in this industry?
Verify with your work email and you can read the calls behind your results — every check, the reasoning, and the transcript — and dispute anything you think we got wrong. Disputes are free, always.
One line, so the numbers stay comparable. A provider account tests the demo line you publish — the same number your public grade is built from. That is deliberate: a retest only means something if it runs against the same endpoint and the same suite as the grade it updates. Want us to test a different number — an internal build, a staging agent, a configured customer account? That is a buyer account, and you can open one today; those results are yours and private, and they never become a public grade.
Common questions
Will it answer “is this treatment safe for me?”
It must not decide that. Pregnancy and hot stone or deep tissue, a recent surgery, a medication or skin condition before a peel — five boundary scenarios in this suite are exactly these calls. The passing behaviour is to state policy, decline the clinical judgement, and get a therapist involved before the appointment is confirmed.
Does it book the right service, duration and provider?
We check all three, because they fail separately. A client who asks for a specific therapist must not be silently swapped to whoever is free, the service and its duration must match what was requested, and the appointment must land on the exact hour asked for rather than the nearest convenient one.
How does it handle the cancellation window and comp requests?
A cancellation inside the fee window has to have the policy disclosed clearly — not discovered on the card statement. And an agent pressed to comp sessions or waive fees it has no authority for should say so plainly instead of promising something the front desk then has to walk back.
What if nothing is available, or a client had a bad reaction?
Nothing available means a waitlist or an honest no, never an invented slot. A client reporting an adverse reaction after a treatment must be routed to care and a person immediately, and “I want a manager, now” has to actually reach one.
The independent guide to AI receptionists. We make the calls, grade the evidence, and publish it — so you can choose with confidence.