Wellness Accounts Get Clinical Messages Without a Clinic

A physiotherapy clinic knows it is in healthcare. It has a registration, a compliance conversation it has already had, and someone who worries about wording.

A yoga studio, a nutritionist, an ayurveda brand, a supplement seller, a sleep coach and a breathwork practitioner mostly do not think of themselves that way. They are wellness. Lighter, friendlier, further from a regulator.

Their inboxes do not know that. The messages arriving are indistinguishable from the ones a clinic gets — people describing what is wrong with them and asking whether the thing you sell will fix it — and they arrive at an account with none of the structure a clinic has around what it is allowed to say.

That gap is the whole subject here. The clinic version of this is about who signs off the wording; the marketing version is about content that invites answers you cannot give. This one is about the accounts that get both problems and think they have neither.

Three things that never go out automatically

An answer to a described symptom. "That sounds like a vitamin D thing" is a clinical judgement produced by a text box. It does not stop being one because the account posts smoothies. A generated reply will produce it fluently and confidently, because the internet is full of confident health content written by people under no obligation to be right.

A claim. How long results take, how much someone will lose, whether it cures, reverses, detoxes, boosts or heals anything, whether it is safe alongside a medication, whether it is safe in pregnancy. Every one of those is a claim your business then owns, in writing, sent to a stranger nobody has assessed.

A price for a programme. Wellness pricing almost always depends on an assessment — how many sessions, which plan, whether there is a concession. A figure sent before that conversation is either a number you revise upward, which reads as bait, or a number you have to honour for a case you would have quoted differently.

You cannot build a detector, so build for failing to detect

The instinct after reading the above is to write a very good negative-keyword list and relax.

Write the list. pain, injury, pregnan, diabet, thyroid, BP, medicine, tablet, dose, depress, anxiety, sleep, child, bacha, cure, safe, and whatever your own inbox actually contains — read a week of it rather than guessing. Matching folds case and punctuation and Latin accents, so Pregnant, pregnant. and PREGNANT are one word. It does not fold Indic vowel signs, and it matches whole words: pain will not catch painful, so write both.

Then accept that the list will miss things. Real distress is often typed short, in lowercase, without any of the vocabulary a keyword list expects, and frequently in a second language or a mix of two. A person writing "kuch theek nahi lag raha" has not used a single word on anybody's list.

So the design cannot depend on catching it. It has to be safe when it does not.

Design the post so the reply is answerable

The real fix is upstream of the inbox, and it costs nothing.

A reel that ends with "comment your biggest health struggle below" has built a machine for generating messages you are not able to handle. A reel that ends with "comment GUIDE for our information sheet" has built one you can run entirely on a free template: send the sheet, offer the callback route, capture the enquiry, stop.

Same content, same reach, completely different inbox on Monday. Wellness accounts almost universally choose the first version because it gets more comments, which is true and is also the problem.

The public half deserves a rule of its own. A warm, specific public reply under somebody's real name tells their followers something about that person's health, even when it contains no clinical detail at all. Say "sent you the details" and move everything else into the DM.

Read what got blocked, because it is your record

Ten checks run in fixed order before any reply is sent, and a blocked reply is recorded with the reason that stopped it. In most businesses that log is housekeeping. Here it is the useful artefact: a written record of what the system wanted to say and what prevented it.

Somebody should be named as the person who reads it weekly. In a solo wellness practice that is you, and it takes ten minutes.

Hold the unusual ones

Most wellness practices are one or two people with low message volume and high individual stakes, which inverts the usual advice. Speed is not the variable worth optimising. Not being wrong is. Takeover helps in the same direction: the moment you reply by hand, automation is out of that thread permanently, with no cheerful template landing under something you wrote carefully.

Cost, and what it will not do

Templated replies are free and unlimited. A credit is spent only when the AI writes a new reply, one credit per reply. An account built this way runs almost entirely on templates, because the entire point is that the wording is fixed, considered and identical for everyone. Free tier is 100 credits with no card; packs start at ₹499. Climbing credit use is a signal worth investigating — it usually means the automation is answering things it should be routing.

It has no calendar, so it cannot book or confirm a session. No CRM sync; leads leave as CSV, and a CSV of people's health complaints is very easy to email to somebody, so decide who may export it. No WhatsApp. It cannot message anyone who has not messaged you first.

Official Graph API only: seven days to answer a comment, counted per comment, and twenty-four hours inside a DM thread, restarted only when they write again.

And it cannot assess anybody's health. Everything above is arranged so it never gets the chance to try. What each of the ten checks does is worth reading before you go live rather than after.