Northway OSSeptember 14, 2026
How to spot a regular who is about to stop coming in
Most owners find out a good customer stopped coming when they happen to think of them. That is usually months after the fact, long past the point where a friendly note would have brought them back.
The usual advice is to pull a list of anyone who has not bought in 90 days. That list is close to useless, and it is worth understanding why before we fix it.
Why a fixed cut-off does not work
Ninety days means completely different things to different customers.
Somebody who came in every Tuesday for a year is in trouble after three weeks of silence. A customer who reliably visits twice a year, before Christmas and before summer, is behaving perfectly normally at day 89. A single rule flags the second person and stays silent about the first, which is exactly backwards.
The problem is that a fixed number treats your customer list as one group. It is not. It is a few hundred individual habits.
Work out each customer's own rhythm
You can do this with nothing more than the sales records you already have. For each customer:
- List their purchase dates over the last 18 to 24 months. You need enough history to see a pattern, and less than a year will mislead you on anything seasonal.
- Work out the gaps between visits, in days.
- Take the typical gap. Use the middle value rather than the average, because one unusual six-month gap will drag an average out of shape while the middle value ignores it.
- Note how consistent they are. Someone whose gaps are 12, 14, 13 and 15 days is a metronome. Someone at 10, 60, 25 and 90 has no rhythm to break.
Now you have a baseline per customer rather than one rule for everybody.
The three signals that come first
Once you have that baseline, three things tend to show up before a customer disappears completely.
They pass 1.5 times their usual gap. This is the single most useful signal. A weekly regular who has not been in for eleven days is drifting. That same eleven days from a monthly customer is nothing. Only the baseline tells you which is which.
Their basket shrinks before it stops. People often trim before they leave. Someone who used to spend $60 a visit and has been spending $28 for the last three visits is usually on the way out, even though they are still technically active.
They stop the extras first. The coffee without the pastry, the cut without the colour, the service without the parts. The add-on is the first thing to go, and it goes weeks before the visits do.
Any one of these is worth a look. Two together is worth a personal note this week.
A worked example
Say a customer's last eight gaps were 13, 15, 12, 16, 14, 13, 15 and 14 days. Their typical gap is 14 days, and they are extremely consistent.
It has now been 25 days. That is nearly 1.8 times their normal rhythm, well past anything in their history. Last year they spent about $1,400 with you.
That is not a marketing list. That is one person, with a number attached, and a reason to reach out today rather than in October.
What to do in the first two weeks
Speed matters more than polish. A note that arrives while they still think of themselves as your customer reads as thoughtful. The same note three months later reads as marketing.
Keep it short, specific and free of offers at first. Mention what they actually buy. "We have the dark roast back in" beats "We miss you" every time, because the first proves you know who they are.
Hold the discount in reserve. If you lead with money off, you teach a good customer to wait for a coupon. Try the personal note first, and keep the offer for people who ignore it.
When not to chase
Not everybody who goes quiet should be chased, and knowing the difference saves you from annoying people.
- One-time buyers. Somebody who bought once, two years ago, was never a regular. They have no rhythm to fall off.
- Discount-only customers. If every one of their purchases came with a coupon, what lapsed is the discount, not the relationship.
- People who told you why. They moved, the project ended, their circumstances changed. Note it and leave them alone.
Doing this without a spreadsheet
All of the above is arithmetic on records you already have. If you have a few hundred customers, it is an afternoon in a spreadsheet, and it goes stale the moment you close it.
Keeping it current is the hard part, and it is the part worth automating. Northway watches these baselines continuously, tells you who has fallen off their own pattern and what they were worth last year, and drafts the note to each one. You read them and decide what sends. Nothing goes out without your approval.
If you want to see what that would surface in your own records, email us and we will take a look together.