Playbooks · 5 min read
Groomers and Boarders: The DM a Bot Must Never Answer
Most of a groomer's inbox is genuinely easy. Do you do double-coated breeds. Kitna time lagta hai. Do you have parking. Do you pick up and drop. Is the boarding air-conditioned. Do you take cats.
All fixed, all asked constantly, all answerable by a free template at eleven at night. That part writes itself.
The part worth your attention is the message that arrives looking exactly like the others and is not like the others at all.
Never let it answer anything about an animal's health
"He's been scratching a lot, is it fleas or allergy?" "Kutte ko ulti ho rahi hai since morning, kya karein?" "Her paw is swollen after the last groom, is that normal?" "Is it okay to bathe a puppy with a slight fever?"
These arrive in a grooming or boarding inbox constantly, and every one of them is a veterinary question wearing the clothes of a service question. The correct behaviour for an automated system is to refuse to answer and get a person involved. Not a softened answer, not a hedged answer with a disclaimer bolted on. No answer.
Two reasons, and neither is abstract. The first is that a generated reply is fluent regardless of whether it is right — it has no way to represent "I have not seen this animal and being wrong here ends badly." The second is that in an animal's case the owner will act on what your account said, immediately, because you are the pet professional they already trust.
This cuts against instinct, because health-adjacent questions are high volume and they feel like exactly the thing a helpful business should answer fast. They are the thing to refuse.
A generated answer about a limping dog is a confident sentence written by something that has never seen an animal.
People send photographs, and the photograph is the message
This is the pet-specific failure that surprises people. A large share of your DMs are pictures. A matted coat, a rash on a belly, a puppy in a carrier, a nail that looks wrong.
A keyword matcher sees an attachment and, at most, whatever caption came with it. "Look at this 😞" contains nothing your negative list can catch, and the automation happily fires a booking template underneath a photograph of an infected ear.
There is no fix inside the trigger. The fix is structural: for the automations most likely to attract photos — anything on grooming, skin, coats, nails — turn off automated DM replies entirely and use the public comment reply to invite them into a thread a person will answer. Slower, and correct.
The disclosures that must reach a person
Boarding and training inboxes carry a second category the automation should never process.
Bite history and aggression. "He's nipped a couple of times, is that okay for boarding?" That is an intake decision affecting your staff and other animals. It needs a person, a phone call, and probably a meeting.
Vaccination and records. Requirements for boarding vary by facility and by local rules, and what is required where you are is not something a template should assert on a stranger's behalf. State your own policy in your own words, ask them to send records to a human, and leave the rest alone.
Anything about an animal that is old, post-operative, or on medication. Route it. Always.
The bookable part, which is most of it
With those carved out, what remains is a large and boringly repetitive volume that is perfect for templates: services offered, breeds you handle, whether you do home visits, drop-off windows, what to bring, whether the day boarding has webcams, how long a full groom takes for a Golden versus a Shih Tzu.
What it will not do is confirm a slot. There is no calendar here — no diary sync, no booking, no hold, no deposit. So the automation names the service, gives the fixed facts, and routes to whoever holds the diary, exactly as a salon should. A confirmation your system cannot verify is a person arriving with a nervous dog to find no appointment.
Quiet hours deserve a thought too. Overnight silence is correct — you do not want your account replying at 2am — but it means an owner writing at 2am about a sick animal gets nothing, which is the right outcome only if the rest of your presence points them at an emergency vet. Put that in your bio, not in a template that fires.
Cost, clocks, and what it does not do
Templated replies are free and unlimited; a credit is spent only when the AI writes a new reply. Since almost all of a pet business's answerable volume is fixed information, the repetitive part costs nothing at any footfall. Free tier is 100 credits, no card; packs start at ₹499.
Official Graph API only: a comment stays answerable for seven days, a DM thread for twenty-four hours from their last message, restarted only when they write again. It cannot book, take payment, message anyone who has not messaged first, or do WhatsApp. Leads leave as a CSV.
If you sell products in this space rather than services, the health line falls in a different place and there is a separate post for it. Otherwise build the refusal list, then one comment-to-DM flow, then read the blocked activity after a week.
