Insights · Sales follow-up

How to notice when a regular customer stops buying a product

Spot lapsed products by each customer's own buying rhythm, not a fixed number of days, and turn the flag into a call and a decision.

To notice when a regular customer stops buying a product, track each customer and product pair separately, measure that pair's usual gap between purchases, and flag it when the gap is clearly overdue. Then have the rep call, record why, and tell purchasing when a line has stopped for good.

The silent loss

A restaurant that has bought your butter every week for a year does not call to say it has switched supplier. It simply stops including butter in its order. The rest of the order keeps coming, so the account looks healthy in every report that counts revenue by customer. The lost product only shows up weeks later, as a dip in a product total that nobody connects to one customer.

This is why "customer churn" is the wrong frame for a food distributor. Customers rarely leave all at once. They leave one product line at a time, and each line has its own rhythm.

Why a fixed number of days fails

The obvious rule is "flag any customer who hasn't bought a product in 30 days." It fails in both directions:

  • A hotel that buys cream twice a week has been gone for a month before the 30-day rule notices.
  • A restaurant that buys a speciality cheese every six weeks triggers a false alarm every single cycle.

The rep learns that the alerts are noise and stops reading them. Any rule that ignores the customer's own rhythm ends the same way.

Define "late" by each customer's own cycle

The better unit is the pair customer × product, which we call a line. For each line, look at the gaps between purchases over the past months and take a typical gap (the median works well, because it ignores the odd holiday). That is the line's own cycle.

Then compare the time since the last purchase with that cycle:

  • Due: the typical gap has passed. A normal reorder is expected now.
  • Late: clearly more than one cycle has passed. Worth a question on the next call or visit.
  • Needs a decision: several cycles have passed. Something has changed, and someone should find out what.

The thresholds are yours to tune. In Cadence, the sales app we built for food distributors, a line that has missed about three of its own cycles comes up for a decision. Lines with too few purchases to have a rhythm are left out until they do.

Turn the flag into a conversation

A list of late lines is only useful if the rep can act on it in a minute. For each line, the rep should see the customer, the product, the usual rhythm, the last purchase date and the last note. The call itself is short: "You usually take the 1 kg unsalted butter every week, and we haven't sent any since the 3rd. Is everything all right with it?"

The answer is almost always one of a few things, and recording which one matters as much as the call:

  • Still buying, order coming: a timing slip. Nothing to change.
  • Switched to another of our products: not a loss. Note it, so the old line stops flagging.
  • Seasonal or menu change: pause the line until a set date.
  • Buying elsewhere: a real loss, often on price or service. Worth a manager's attention.
  • Stopped for good: close the line, take it out of the sales plan, and tell purchasing.

Close the loop with purchasing

The last answer is the one most companies lose. When a rep learns that a customer has stopped a product, purchasing should hear about it the same day, not discover it later as slow-moving or short-dated stock. If your sales plan is built per customer and product, a stopped line simply drops out of the plan, and the purchasing view adjusts.

Make it a weekly routine

A system that flags lines is only half of it. The other half is a short weekly habit:

  1. Each rep clears their "needs a decision" lines before the weekly sales meeting.
  2. The manager reviews only the lines marked "buying elsewhere" and asks why.
  3. Once a month, look at which products appear most often in that list. That is a pricing or quality conversation, not a sales one.

Kept up for a quarter, this turns a vague feeling that "some customers are buying less" into a named list of lines, reasons and decisions.

Can you do this in Excel?

Partly. With a sales export by customer, product and date, a spreadsheet can calculate each line's typical gap and days since the last purchase. Where Excel struggles is the rest: putting the right lines in front of the right rep every morning, recording the answer, and making sure a stopped line reaches purchasing. That is where a small sales app, built on your ERP export, earns its place.

Where this fits: this is the method behind Cadence, our sales app for food distributors. Also see software for food and beverage distributors.

FAQ

Questions readers ask

How much sales history do we need?

Enough for each customer and product to have several purchases. For weekly or fortnightly lines, a few months is usually enough; slow lines need longer and are better left out until they have a rhythm.

Isn't this just churn analysis?

Churn looks at customers. In food distribution, customers usually leave one product at a time, so the useful unit is the customer and product pair.

Who should make the call, the rep or customer service?

The rep who owns the account, because the answer often changes the next visit. Customer service can help with lines that only need a reminder.

Do we need to change ERP to do this?

No. It works from the sales export most ERPs already produce. Cadence, for example, reads that export and does not write to the ERP.

Contact

See your own late lines.

Send us a sales export and we'll show you which customer and product lines would be flagged this week, on a 30-minute call.

We reply within one working day, Bangkok time (GMT+7).

30 minutes, no deck.

Book a 30-min call