Lapsed Customer Definition: How to Know When a Buyer Has Really Gone Quiet

You have a list of past customers who have not ordered in a while, and you want to send them a win-back flow. The question is where to draw the line: 90 days, six months, a year? None of those is right unless your own order data says so, and the answer is usually different for every product you sell.

Key takeaways

• A lapsed customer is someone who has gone well past their normal reorder window for the product they buy, not someone who crossed a calendar mark like 90 or 180 days.

• Define the window per product from your own order history: the median gap between orders marks "on time", and the 90th percentile gap marks "lapsed".

• Use four states: active, late, lapsed and churned. Each one gets a different action, from nothing to suppression.

• Lapsed first-time buyers, who never placed a second order, are a different problem from lapsed repeat buyers and need their own window and their own message.

• You can build the whole segment from an order export with customer id, order date and product.

What is a lapsed customer?

A lapsed customer is someone whose time since their last order is well beyond the normal gap between orders for what they buy. They are overdue, but not so far gone that you should stop trying.

Many stores use one fixed rule, such as "no order in 90 days". That works only if everything you sell is bought on roughly the same rhythm. A customer who buys coffee every month is already overdue at 45 days. A customer who buys a backpack once a year is perfectly on schedule at 90 days.

Lapsed customers are still a warm audience: they already trusted you once, and their purchase history tells you what they need next. That usually makes them easier to bring back than new customers are to find. Get the definition wrong and you either chase people who were never late, or reach real lapsers after they have found another store.

How do I define a lapsed customer for my store?

The lapsed customer definition that works is one built from your own customer data. Measure the typical number of days between orders for each product, then call a customer lapsed once they are further out than almost all of your repeat buyers ever get. In practice that means the 90th percentile of reorder gaps, per product.

The method:

1. Pull your order export with customer id, order date and product.

2. For each product or product family, take every customer who bought it more than once and compute the days between consecutive orders of it.

3. Take the median of those gaps. That is the "on time" marker.

4. Take the 90th percentile of the same gaps. That is the long-stop: nine out of ten reorders happen before it.

5. A customer past the median is late. A customer past the long-stop is lapsed.

Use the median rather than the average. A handful of customers who come back after a year drag an average far to the right, as the purchase frequency guide explains. If you already track reorder rate per product, the same gap data feeds both numbers.

When is a customer churned rather than considered lapsed?

Churned customers are the ones you have stopped expecting back. They are well past the long-stop and have ignored your win-back sequence, so regular campaigns to them do more harm than good.

The progression runs active, late, lapsed, churned. Where you put the churn line is a business choice, not a statistic. A common approach is to set a second cut-off comfortably beyond the long-stop and move anyone who crosses it, with no reaction to the win-back emails, into the churned group. The ecommerce churn rate article covers how to turn that into a monthly rate.

Take churned customers out of your regular promotions. Old, cold contacts who receive every campaign hurt your sender reputation and inflate unsubscribes. Keep these dormant customers in your records and contact former customers rarely, for a genuine product change or a new line.

How is a lapsed first-time buyer different from a lapsed repeat buyer?

A lapsed first-time buyer never placed a second order. A lapsed repeat buyer used to come back and then stopped. They need different windows and different messages.

For first-time buyers, the window comes from the gap between first and second order, measured only on customers who did place a second order. That gap is often longer and more spread out than the gap between later orders, because the first order is when people are still deciding whether the product works for them. The second order is the hardest one to earn, and customers who make it tend to keep going.

For repeat buyers, the window comes from the gaps between later orders. A regular who misses their usual reorder has changed something: they ran into a problem, found a cheaper option, or stopped needing the product. Use customer feedback to understand why customers lapse before you offer an incentive; the answer also helps with preventing future lapses.

Segment lapsed customers into these two groups, with separate lists and separate flows. A lapsed first-time buyer needs a reason to try again. A lapsed regular needs to be asked what changed.

How do I re-engage lapsed customers once I have the list?

Sort them, contact the ones worth contacting with a message tied to what the customer bought, and leave the rest alone. Many businesses run re-engagement campaigns because bringing lapsed customers back is often more efficient than replacing them. Re-engaging lapsed customers works best when each message has one clear job.

Prioritize by product and value. Start with lapsed first-time buyers of replenishable products, then lapsed repeat buyers with a high past order frequency or average order value. RFM scores are a quick way to rank them; the RFM analysis glossary entry explains the three scores.

Match the message to the product. Targeted messages should use past behavior and previous interactions to shape personalized offers. For consumables, a restock reminder with the exact product and a direct reorder link. For products that do not run out, the accessory, refill or companion item that other customers with the same past purchases bought next. Customers who interacted with your brand recently or showed interest in a product respond better when the outreach reflects it. Email is the default channel; direct mail can make sense for high-value former regulars.

Hold back on discounts. A discount to every lapsed customer teaches your punctual customers to wait for one. A discounted rate can motivate lapsed customers when used selectively, and higher-value segments can get early access or exclusive access to a new product instead. If you test an incentive, such as free shipping on the next purchase, give it only to the lapsed group, with a hold-out that gets no offer, so you can see whether it created orders or just discounted ones that would have come anyway.

Measure what comes back. Set key performance indicators before launch, such as reactivation rate, repeat purchases and click through rates. Track how many reactivated customers place an order within one normal reorder window of the message, against the hold-out. Reactivated customers are a customer segment worth watching on their own: if they lapse again at the same point, the problem is the product or the customer experience, not the timing. The win-back email timing guide covers the sequence itself.

Active, late, lapsed, churned: what each state means

Four states give you a clearer plan than "active or lost".

• Active: the last order is inside the median gap for the product. No retention action beyond normal marketing.

• Late: past the median but inside the long-stop. A friendly reminder in a timed replenishment email belongs here, before the customer ever counts as lapsed.

• Lapsed: past the long-stop. This is where the win-back sequence starts.

• Churned: past your churn cut-off with no response to win-back. Suppress from regular re-engagement campaigns.

RFM recency scores map roughly onto these states, but recency alone treats a coffee buyer and a backpack buyer the same way.

Why one store-wide cut-off fails for mixed catalogues

A single rule forces a bad trade. Set it short and you nag customers of slow-moving products who are not late at all. Set it long and customers of fast-moving products are gone before you notice.

Coffee, supplements and pet food get reordered on short cycles. Furniture, luggage and jewelry may be bought once a year or less. Each product family needs its own median and its own long-stop. For a customer who buys from several families, decide which product defines their state: usually the most frequently reordered one they buy.

Customers who only ever bought a one-off item, such as a gift or a piece of furniture, do not belong in a lapsed sequence at all.

Worked example: one customer, two products

An illustration with round numbers, not a benchmark.

A store sells coffee pods and a premium backpack. From its order history:

• Coffee pods: median gap between repeat orders 30 days, long-stop (90th percentile) 45 days.

• Backpack: median gap between related purchases 365 days, long-stop 540 days.

A customer last bought coffee pods on May 1 and the backpack on January 1. Today is June 16.

This customer is lapsed for coffee and on schedule for the backpack. The outreach should be about coffee: a restock reminder with a reorder link, not a general "we miss you" email.

A fixed 90-day rule gets both wrong. It would not flag the coffee lapse for another 44 days, by which time the customer has almost certainly bought coffee somewhere else, and it would already have flagged the backpack buyer as lost when they are months away from needing anything.

How to identify lapsed customers from your own customer data

A CSV export from Shopify, WooCommerce or most other platforms is enough; the same data can also come from a webhook or API connection to your store.

Columns you need: customer id or email, order id, order date, product or SKU.

Steps:

1. For each customer and product, list the orders in date order.

2. Compute the days between consecutive orders of that product, across every customer who bought it at least twice.

3. For each product, compute the median and the 90th percentile of those gaps. Skip products with too few repeat customers to trust; a few dozen is a sensible floor.

4. For each customer's latest order of each product, compute the days since that order and compare it with the median and the long-stop.

5. Give each customer one state, based on their most frequently reordered product.

6. Export three lists: late (reminder), lapsed (win-back) and churned (suppress). These lists are what your retention and customer success teams work from.

A spreadsheet with pivot tables handles a small catalogue. For many products or several stores, Affinsy reads the same export, computes the reorder cadence per product and the RFM segments, and exports the customer lists as audiences you can load into your email tool, or upload as custom audiences in ad platforms.

Next steps for customer retention

• Export your orders and compute the median and 90th percentile reorder gap for your top products.

• Tag every customer as active, late, lapsed or churned.

• Put a reminder on the late list and a win-back sequence on the lapsed list, with a hold-out group, and let automated sequences run them once the definitions are set.

• Write the definitions down so reports, flows and your wider marketing strategy use the same windows.

• Recalculate the windows each quarter, and after a new product line or a big promotion.

If you would rather have the reorder windows computed for every product and the late and lapsed lists ready to export, the 48-hour analysis delivers them from your own orders.

FAQ

How often should I recalculate lapsed customer windows?

Quarterly works for most stores. Recalculate sooner after a new core product, a pack-size change, a price change or a large promotion, since each can shift how long customers take to reorder.

How do subscriptions fit into the definition?

Subscribers have an expected shipping date, so the definition is simpler: a subscriber who cancelled or failed payment and did not restart within one or two normal cycles is lapsed. Keep them in a separate flow from lapsed one-time buyers, because the reason and the fix are different.

Should email engagement change whether a customer counts as lapsed?

Keep the definition based on orders, which reflect customer behavior most directly. Use customer engagement signals such as opens, clicks and browsing history to judge engagement level and decide how hard to push and on which channel, so you avoid over communication. Do not use them to decide who is lapsed.

Do I need loyalty programs to win back lapsed buyers?

No. A reminder timed to the product and the right next item do more for most lapsed buyers than points. For customers with a high lifetime value, a loyalty perk can help rebuild the entire relationship rather than drive one extra order, and some of those returning buyers become loyal advocates. Loyalty programs can reward loyal customers and existing customers later, once you know who comes back on their own.

What if I do not have enough orders to compute percentiles?

Start with a reasonable guess from how long one unit lasts in normal use, and label it as a guess. Record reorder gaps from now on, and replace the guess with the measured median and long-stop once enough customers have reordered.