Learn · customer-understanding

By Fieldwork Research Team · Published 2026-07-16

Why don't customers come back?

Customers who don't return after their first purchase are the most expensive problem in most businesses. The acquisition cost is sunk, the relationship was never profitable, and the customer is now available to a competitor at no acquisition cost to them. Most businesses know their repeat purchase rate. Very few know why it is what it is, because the customers who don't come back simply disappear. They don't complain. They don't unsubscribe. They just stop.

Why repeat purchase is the metric that determines profitability

Acquiring a new customer costs between 5 and 25 times more than retaining an existing one, depending on the industry. A business where most customers only buy once is a business that spends most of its marketing budget replacing customers it already had. The unit economics only work at scale, and they only work at scale if the cost of acquisition stays below the lifetime value: which, for a one-purchase customer, is a single transaction.

The businesses with the best unit economics are the ones where customers come back. Not because they were incentivised to with discounts and loyalty points, but because the first experience was good enough that returning felt like the obvious thing to do.

Understanding why customers don't return is therefore not a retention problem in isolation. It is the central question of whether the business model works.

What the data tells you about customers who don't return

Cohort analysis will show you that a certain percentage of customers make a second purchase within 90 days, and what that percentage is for different acquisition channels, product categories, and average order values. This is useful structural information.

It doesn't tell you what the experience was like for the customer who bought once and never came back. It doesn't tell you whether they had a problem with the product that they didn't bother to raise. It doesn't tell you whether they simply forgot you existed. It doesn't tell you whether they found an alternative that was marginally better for their specific situation.

Each of these has a different solution. Forgetting you existed is a remarketing problem. A product quality issue is a product problem. A marginally better alternative is a competitive positioning problem. Making the wrong diagnosis leads to the wrong investment.

The most common reasons customers don't return

Talking to customers who bought once and didn't come back reveals patterns that repeat across different categories and price points.

The first experience was good enough but not memorable. The product did what it said it would. The delivery was fine. Nothing went wrong. But nothing was remarkable either. When the customer next needed something similar, they didn't have a strong reason to return specifically to this business rather than searching fresh. Satisfactory is not sufficient for repeat purchase.

This is the most common pattern and the most difficult to address, because there's no specific failure to fix. The fix is making the first experience actively memorable, not just adequate.

Something about the post-purchase experience created friction. Not the product itself but the experience around it. The returns process was more complicated than expected. A query to customer service took longer than felt reasonable. The packaging was harder to deal with than anticipated. The account portal didn't work on mobile. Small frictions that didn't prevent the first purchase do prevent the second, because by the second purchase the novelty has worn off and the friction feels larger.

The product didn't perform as well as expected over time. The customer's first-use experience was positive, but over the following weeks something fell short of what they'd expected. The product wore out faster than anticipated. The service didn't hold up. The initial experience was better than the ongoing one.

They found a marginally better alternative without looking for one. An ad appeared. A friend recommended something. A competitor's product showed up in a relevant context. The customer didn't seek out an alternative: one was presented to them in a moment when they were open to it. And the first experience, while fine, wasn't strong enough to create resistance.

They simply forgot. For purchases that are infrequent or not tied to a recurring need, the customer completed the purchase, the transaction closed, and the business left no ongoing presence in their life. When the need arose again, they had no particular reason to think of this business first.

How to find out which pattern is operating

The window for useful data is narrow. A customer who bought eight months ago and hasn't returned has a fuzzy memory of the experience and may not be able to give you useful specifics. A customer who bought six to eight weeks ago and hasn't made a second purchase is the most informative: recent enough to remember the experience clearly, long enough past the purchase that the decision not to return is established rather than just delayed.

The conversation should cover:

The purchase decision itself. What were they trying to solve? What made them choose this business specifically? Was it their first purchase from this category or were they switching from somewhere else?

The experience after purchase. Not a satisfaction rating: a story. Walk me through what happened after you received the order. What did you use it for? Was there anything that surprised you, positively or negatively? Did you have any reason to contact us?

What they've done since. Have they bought from a similar business since then? What drove that decision? If they haven't needed the product category again, what would prompt them to?

The direct question, asked last. If you were going to buy something similar again, would you come back here, or would you search fresh? Why?

The answer to that last question, and the reasoning behind it, is the most direct signal available about what's preventing repeat purchase.

What this looks like in practice

A specialty food retailer has a first-purchase repeat rate of 22% within 90 days: meaning 78% of customers who buy once don't buy again within three months. The team assumes the primary issue is that customers forget the brand exists and invests in a retargeting campaign.

The retargeting campaign improves the repeat rate to 24% at significant cost. The team considers this a reasonable result.

A researcher interviews 15 customers who bought once and didn't return within 90 days. The interviews reveal that the retargeting assumption was correct for about a third of customers. They did forget, and a retargeting ad did remind them, and some of them came back. But for the majority, the pattern was different.

Nine of the fifteen customers described a post-purchase experience that left them with an unresolved question. Seven had received an order that was slightly different from what they expected: a different size, a different packaging format, and didn't know whether this was standard or an error. None of them contacted customer service. They just noted it and moved on. When asked whether it affected their likelihood of returning, six of the seven said they'd probably search around next time rather than assuming the same thing would happen.

The fix is a proactive post-purchase message that explains the packaging variation, invites questions, and makes it easy to get an answer. That message costs almost nothing to send and addresses the root cause that the retargeting campaign was compensating for without solving.

The repeat purchase rate moves to 31% in the following quarter after the message goes live. The retargeting campaign is scaled back.

Frequently asked questions

How soon after a first purchase should I try to interview customers who haven't returned?

Six to ten weeks after purchase is the optimal window for most businesses. Early enough that the experience is fresh and specific, late enough that the absence of a second purchase is a real signal rather than just timing. For businesses with long purchase cycles (furniture, cars, annual software subscriptions), extend this window to reflect the natural repurchase timeline.

Is it worth interviewing customers who only made one purchase a long time ago?

For recent enough purchases (within six months), yes. Beyond that, the specificity of recall drops significantly and the customer's situation may have changed enough that their experience is no longer representative of what current customers are experiencing. The most useful interviews are with customers whose experience was recent and whose non-return is recent.

How do I distinguish between customers who didn't return because of a problem and customers who simply haven't needed the product again?

Ask directly in the screening process. "Have you bought [product category] from any business since your purchase with us?" If the answer is no and their purchase cycle is naturally long, this customer's non-return may not be a problem to solve. If they say yes. They bought the same product category elsewhere. That's the conversation to have.

Should I offer a discount to get lapsed customers to return?

Discounts reactivate price-sensitive customers. They don't address the underlying reason for non-return. A customer who had a confusing post-purchase experience won't come back because of a 15% discount. They'll come back, have the same confusing experience, and leave again. Understanding why customers don't return before trying to bring them back produces better long-term results than incentivised reactivation.

What response rate should I expect when reaching out to lapsed customers for interviews?

Lower than for current customers: typically 8 to 15% for a personal, specific outreach. An incentive (store credit, small gift card) helps. The framing matters: a message that expresses genuine curiosity about their experience converts better than one that feels like a retention play. Customers who feel that the business actually wants to understand what happened are more likely to give their time.

How is this different from a standard NPS survey?

NPS asks customers to rate their likelihood to recommend on a scale of 0 to 10. It produces a score that tells you whether sentiment is positive or negative, and tracks that score over time. It does not tell you what the specific experience was that produced the score, what would have changed it, or what the specific barrier to returning is. Customer interviews go where NPS cannot: into the specific experience, the specific moment things went wrong, and the specific change that would bring someone back.

Related on Fieldwork

Last updated: 2026-07-16