How to win back customers who bought once and left
Seven ready segments — expired subscribers, abandoned checkouts, dormant buyers. Free gets one campaign ever, Pro three a week. Why the cap is the point.
TL;DR. Every shop accumulates people who bought once and vanished, subscriptions that lapsed, checkouts that got to the payment screen and stopped. Win-back campaigns write to exactly those groups: seven ready segments, narrowed by recency, with media and buttons. Anyone who opted out of marketing is removed before you ever see the count. The free plan allows one campaign ever and Pro allows three a week — and that cap is not a paywall dressed as a limit, it is the thing that stops the feature from turning your customer list into a burned list.
The people most likely to buy from you again are the ones who already bought once. That is the least surprising fact in commerce, and the reason it stays theoretical for most sellers is that finding those people is work: who lapsed, who almost paid, who bought in March and hasn’t opened the shop since.
The seven segments
Four are about your channel. Expired subscribers are the clearest case — they paid, the subscription ran out, they didn’t renew. Abandoned checkouts are people who reached the payment step and didn’t finish. Everyone currently subscribed is there for announcements rather than win-back. People who left are the hardest audience and the one to treat most carefully.
Three are about your shop. Everyone who bought. Buyers of one specific product — for a restock, a new version, a related thing. And buyers who have gone quiet, which is the closest thing to a classic win-back list.
Each can be narrowed by how recently the thing happened, and this matters more than picking the right segment. “Bought and went quiet” is a different message at three weeks than at nine months.
What abandoned actually means
A checkout counts as abandoned when there is an expired payment attempt and no active subscription.
Both halves do work. Someone who bounced off the payment screen and came back an hour later to pay is not an abandoned checkout, and writing to them as if they were is the kind of message that makes a customer trust you less than before you sent it. The second condition is what keeps the segment honest.
The cap, and why it’s the useful part
One campaign, ever, on the free plan. Three per seven days on Pro.
It is tempting to read that as a paywall, and on the free plan it partly is. But the weekly cap on Pro is not about money — three a week is far below what a determined marketer would send, and that is the intent. A list of past customers is a resource you can spend exactly once: message it too often and the opt-outs and blocks are permanent, and no plan tier gets them back.
So the practical shape of this feature is not “how many can I send” but “is this one worth sending”. If you are on the free plan, that question has a very sharp edge: you get one, and there is no second.
What you cannot un-send
Everyone who opted out of marketing is removed from the audience before the campaign is built. The count you see on screen has already had them taken out, which means the number is smaller than your customer list and that difference is the honest one.
Opting out is one tap for the customer and permanent for your workspace. That is the right design and it is also the reason the cap exists: every campaign that lands badly makes the next audience smaller, and you never find out which message did it.
Write it like a message, not a broadcast
A campaign carries media and inline buttons, the same ones a channel post can carry. Use the button. A win-back message without one is asking someone who already stopped caring to go and find the thing themselves.
The delivery list is worth opening once after each send. The gap between audience size and delivered count is people who have blocked the bot — and that gap growing between campaigns is the clearest signal you have that the previous one was a mistake.
If… then…
| If… | Then… |
|---|---|
| Subscriptions are lapsing | Expired subscribers, narrowed to the last few weeks |
| People stall at the payment screen | Abandoned checkouts — they wanted it |
| A product is back in stock | Buyers of that product, nobody else |
| Your shop has gone quiet | Dormant buyers, and give them a reason, not a reminder |
| You’re on the free plan | You have one. Spend it on the segment most likely to convert |
| You want a weekly newsletter | This is the wrong tool — the cap is there to prevent that |
| The delivered count keeps dropping | Stop and rewrite; you are burning the list |
Send it to yourself first
A campaign can be test-sent to you before it goes anywhere — the real message, with the real footer, buttons and media, delivered to your own chat. It does not touch the quota: a draft only counts once it is scheduled or sent.
On the free plan this is not a nicety, it is the whole safety net. You have one campaign ever; the test costs nothing and is the only chance you get to read the thing as a customer would.
Where to start
Look at expired subscribers first, narrowed to the last month or two. They are the segment with the least explaining to do: they already decided you were worth paying for once, and the lapse is often nothing more than a card that failed.
Write one message with one button and one reason to come back that isn’t “we miss you”. Then open the delivery list and note the number, because that number is your baseline for whether the next one was worth sending.
And if you are on the free plan, do the arithmetic before you press anything: one campaign, no second chance, and the segment you pick is the whole of the decision.
The shop these buyers came from — the shop page. Buttons that go in the message — add buttons to a Telegram post. Selling in the first place — how to sell on Telegram.
The other end of the same lifecycle — Telegram welcome message: a sequence, not one greeting.
What people usually ask
- Who exactly can I write to?
- Seven ready segments, not a free-for-all. On the channel side: subscribers whose subscription expired, people who started a payment and never finished, everyone currently subscribed, and people who left. On the shop side: everyone who bought, buyers of one specific product, and buyers who have gone quiet. Each can be narrowed by how recently the thing happened.
- What counts as an abandoned checkout?
- Someone who has an expired payment attempt and no active subscription. Both halves matter: an expired attempt alone isn't abandonment if they came back and paid later, and this is the difference between a useful nudge and an insulting one.
- How many campaigns can I send?
- On Pro, three in any seven-day window. On the free plan, one — not one a month, one ever. The cap is deliberate and it is the most useful thing about the feature.
- Will people who unsubscribed from marketing get it?
- No. Anyone who opted out is excluded from the audience before the campaign is built, so the count you see already has them removed. Opting out is a single tap for the customer and it is permanent for that workspace.
- Can I put a button in the message?
- Yes — the same inline buttons a channel post can carry, plus media. A win-back message with no button is a message asking someone to go find the thing themselves.
- Can I see who actually received it?
- Yes, each campaign keeps a delivery list. Worth reading once: the gap between audience size and delivered count is people who blocked the bot, and that number tells you something about the last campaign.
- Can I see it before it goes out?
- Yes — a test send delivers the real message to your own chat, footer and buttons included. It does not count against your quota, because a campaign only counts once it is scheduled or sent. On the free plan, where you have exactly one campaign, this is the only rehearsal available.