Two posts on this site are about keeping clients: where your clinic's clients actually go diagnoses why they leave, and the follow-through system is the routine that stops it. This one is about the people that arrived too late for — the clients who stopped coming a year ago and whom nobody has contacted since.
It is a genuinely different job. Retention is about a relationship that still exists. Win-back is about one that lapsed, possibly for a reason, possibly involving something that went wrong that you never heard about. The tone that works is completely different, and the biggest risk is not being ignored — it is sending something clumsy to someone whose animal died.
First, build the list properly
The definition matters more than the mechanics. A useful lapsed segment for a small-animal clinic is something like: an owner with a pet on file, at least two prior visits, and no visit in the last twelve months. The two-visit floor is doing important work — a single-visit client was probably passing through or trying you out, and treating them as a lapsed regular misreads the relationship.
Then, before you send anything, take things out of the list. This is the step that separates a win-back from an embarrassment.
- Anything marked deceased. Obvious, and it still goes wrong constantly, because the record was updated in a consult note and not in the field the reminder query reads. Check the free-text notes on this segment specifically before you send.
- Very old animals with no recent visit. A seventeen-year-old cat last seen fourteen months ago is a conversation to have carefully or not at all. If the record does not say, err towards not sending.
- Anyone who has opted out of messaging. If they said no once, no is the answer, and re-contacting them is both rude and a consent problem.
- Anyone with an open complaint or an unresolved billing dispute. A cheerful "we miss you" into that situation makes it worse. Handle it as a phone call from a human or leave it.
Doing this properly takes an hour on a list of a few hundred and it is the hour that determines whether the campaign is worth running. If your records cannot support these exclusions, that is the finding — fix the record before you run the campaign.
What to send, and what makes it land
The instinct is a discount. Resist it, at least in the first message. A discount frames the relationship as transactional and it trains the segment to wait for offers, but more practically it answers a question nobody asked. The reason most lapsed clients have not come back is not price. It is that nothing prompted them and the animal seemed fine.
What works better is a specific, useful, slightly overdue reason to come in. You have one in the record: something is due. A vaccination, a deworming, a senior check, a dental that was mentioned and never happened. Lead with that.
Compare two messages. "We miss you at [clinic]! Come back for 20% off any service this month." Against: "Hello — this is Dr Rao's clinic. Our records show Simba's annual booster was due in March and we have not seen him since. Everything all right? If he is due a check we have Thursday evening or Saturday morning free." The second names the animal, gives a real reason, opens a door for the owner to tell you the animal died or moved, and proposes a specific slot. It also reads as though a person sent it, because a person could have.
That question — "everything all right?" — is doing more work than it looks. A meaningful share of your lapsed list is not winnable, and the fastest way to find out is to ask in a way that makes it easy to say so. An owner who replies "he passed away last year" has given you a record correction and closed the loop kindly; an owner who replies "we moved to Pune" has done the same. Both are better outcomes than silence.
Why WhatsApp specifically, and the rule it imposes
The whole approach depends on the message being repliable. An SMS blast into an Indian inbox produces no conversation, which means the exclusions you could not make from your records never get made and the owner whose animal died has no easy way to tell you. On WhatsApp people reply, which is the entire mechanism here.
One practical constraint to plan around: for clinic-initiated messages outside an active conversation, the WhatsApp Business API requires pre-approved templates. Your win-back opener is exactly that kind of message, so the template has to exist before the campaign does. Write it, get it approved, then run the list. Sending prescriptions and invoices over WhatsApp covers the app-versus-API decision and the template mechanics.
The other thing to check before you start is how your software bills messages. If you are charged per message, a few hundred win-back messages is a line item, and the incentive quietly pushes you towards a smaller list — which is backwards, since this is the cheapest revenue available to a clinic that already exists. CliniCore includes WhatsApp messaging in the plan fee rather than metering it, for exactly that reason.
Run it small, and run it as a conversation
Do not send to four hundred people on day one. Send to thirty, read every reply, and find out what your list is actually made of — you will almost certainly discover that your record quality is worse than you thought, and it is much better to discover that on thirty than on four hundred.
Whoever sends it must be able to answer the replies within the hour, in a real voice, and to book on the spot. A win-back that generates twenty replies and answers them on Thursday has converted a warm response into a second abandonment. If nobody can staff the replies this week, run it next week instead.
And send once. One message, one follow-up at most for people who did not reply, then stop and update the record. A lapsed client who is chased is not a lapsed client, they are an annoyed former client, and you will need that phone number to still work in two years when their next animal is unwell.
What to measure, and what it tells you
Two numbers. Reply rate tells you whether your message and channel work. Appointments booked within thirty days tells you whether the campaign was worth running. Everything else — opens, deliveries, sends — is activity, not outcome.
The third number is the one nobody expects to care about and usually matters most: how many records you corrected. Animals recorded as alive who are not, owners who moved, numbers that no longer work. A win-back campaign is the cheapest data-cleaning exercise available to a clinic, and a clean list makes every reminder you send for the next five years slightly more effective.
If the exercise turns up a lapsed list that is much larger than you expected, that is a retention problem rather than a win-back one, and the fix is upstream: the follow-through system is where to go next. The free 60-second audit will also put a rough number on what the recall gap is costing, without a signup.