How to win back lost dispensary customers

Define lapse from your data, not from a calendar

A lapsed customer is not someone who has not visited in 90 days. It is someone who has not visited in noticeably longer than their own normal gap. A flower buyer who came in every nine days and has been gone for 25 is a serious problem. A twice-a-year cartridge buyer at 60 days is behaving normally.

Pull your purchase intervals by category and find the median gap. Flower typically sits somewhere around one to two weeks, vape carts three to six weeks, edibles longer and lumpier. Your lapse threshold should be roughly double the median for that customer's dominant category.

Doing it this way changes who you contact. Instead of one enormous 90-day list, you get a rolling weekly set of people who are genuinely drifting right now — which is exactly when a message still works.

Why customers actually leave

Very few leave angry. The common reasons are boring: they went somewhere closer that week, the item they liked was out of stock, someone else texted them a reason to shop, or life simply happened and the habit broke.

Habits break faster than loyalty does. That is the good news — the relationship is usually still intact for a while, and a single relevant reminder can restart it. It is also the bad news, because the window is short. Reach a drifting customer at two weeks past their normal gap and you often get them back. Reach them at six months and you are acquiring a stranger.

This is why detection matters more than the offer. Most stores have a perfectly good coupon and no system that notices who needs it.

The three-message win-back sequence

Message one, sent right at the lapse threshold, should not be a discount. Lead with what is new in the category they buy, plus a light nudge. A meaningful share of recovery happens here, at zero margin cost.

Message two, roughly 7 to 10 days later if there is still no visit, carries the real offer. Make it specific and genuinely worth the trip — a dollar amount off their category, a BOGO on something relevant, or a bundle. Vague 'we miss you' texts underperform badly.

Message three, another 10 to 14 days out, is the last attempt: a short deadline and a clear reason to come in this week. If they do not respond, stop and move them to a low-frequency segment. Continuing to message a non-responder costs money, raises opt-outs, and hurts the deliverability of every other campaign you send.

Keep each message under two lines, include one action, and use a unique code so the register can confirm the recovery.

Offers that recover without training discount-seekers

Anchor the offer to the category the customer actually buys. A flower buyer does not care about 20% off edibles, and the mismatch reads as a mass blast.

Prefer a defined dollar value with a deadline over an open-ended percentage. It converts better and it ends, which prevents customers from learning to wait for the next one.

Suppress anyone who visited since the sequence started, anyone already in another campaign that week, and anyone who has ignored three consecutive win-back attempts. Suppression is what keeps this profitable at scale, and it is the part manual processes always get wrong.

Automate it so it happens every week

Win-backs fail in most stores for one reason: nobody has time to rebuild the list every Monday. Cannabis Text Marketing solves that structurally. It reads purchase cadence from your POS, flags each customer against their own normal interval, and fires the sequence automatically as they cross the threshold — so recovery happens continuously instead of whenever someone remembers.

It also handles the parts that protect the channel: suppression of recent buyers and non-responders, quiet hours, consent and opt-out handling, and managed deliverability so the message actually lands.

At $0.025 per text with no platform fee or contract, a weekly win-back to a few hundred drifting customers costs a few dollars and routinely returns hundreds. Create your account, connect your store, and let it start catching them for you.

What to do with the ones who do not come back

A properly run win-back sequence still fails on most of the people it reaches, and that is normal. What you do with the non-responders determines whether the program stays profitable.

After three ignored attempts, stop the sequence. Move that customer into a low-frequency segment that receives a few messages a year at most — a major holiday event, a genuine store change, a new location. Continuing to message someone who has ignored you three times costs money on every send, raises your opt-out rate, and teaches carriers your traffic is low quality, which damages delivery for the customers who do respond.

Also separate the genuinely gone from the never-really-there. Someone with a single purchase who lapsed was never a regular; a one-message attempt is enough. Someone with eight visits over five months who vanished is worth the full sequence and, if it fails, worth a personal follow-up from a manager if your market and rules allow it.

Finally, close the loop upstream. Look at why customers lapse in the first place. If your win-back lists are consistently full of people who bought one specific product, you may have a stock-out problem rather than a marketing problem. Recovery campaigns are far cheaper than acquisition, but preventing the lapse — with a welcome flow and a replenishment nudge on each customer's own cadence — is cheaper still.

Set a realistic bar for success. Recovering 5% to 12% of a rolling lapsed pool per cycle is a healthy result, and even the low end usually returns many times the cost of the messages at $0.025 each. The stores that do best are not the ones with the cleverest copy — they are the ones whose system catches every drifting customer, every week, without anyone remembering to look.