Treat every late shipment, slow refund and unanswered message as a lifetime-value event, not an order event. Count leaks per 100 orders, measure how long each one took to notice and fix, and fix the causes that repeat.
Most shops judge a problem by the order it touched. A package that arrives two days late costs a little goodwill. A refund that takes nine days costs a support ticket. A question nobody answers costs nothing at all, as far as the dashboard can tell.
That accounting is wrong in a way that compounds. The cost of a bad experience is rarely the order it happened on. It is the orders that buyer would have placed next, and the reviews other buyers read before deciding whether to place their first.
We call these misses CX leaks: small, operational failures after the sale that quietly drain customer lifetime value. This piece looks at what recent research says about how often they happen, what buyers do afterwards, and how to find them while they are still cheap to fix.
What counts as a CX leak
A CX leak is any point after checkout where the buyer’s experience falls short of what they were promised or reasonably expected. On a marketplace like TikTok Shop, most of them are operational rather than about the product itself.
| Leak | What the buyer experiences |
|---|---|
| Late dispatch | The order sits unshipped past the window they were shown. |
| Stalled tracking | The package stops moving and nobody explains why. |
| Slow or disputed refund | They return the item and wait, or have to chase the money. |
| Unanswered message | A question about size, delivery or a fault goes without a reply. |
| Listing mismatch | The item differs from the photos, size chart or description. |
| Stockout cancellation | An order is cancelled after purchase because stock ran out. |
None of these is dramatic on its own. That is what makes them leaks rather than outages: they happen in ones and twos every day, and each one looks too small to deserve attention.
How often experiences go wrong
Qualtrics XM Institute’s research for its 2026 consumer trends found that about 11% of consumer experiences are bad, roughly one in nine. Of those bad experiences, 47% lead the customer to cut their spending, and in online retail the figure is 58%, among the highest of any industry it measured. Qualtrics estimates about $3 trillion in global sales are at risk from bad experiences in 2026. [1] Qualtrics XM Institute — $3 trillion is at risk due to bad customer experiences in 2026
One in nine is the number to hold onto. A shop shipping a few thousand orders a month is not dealing with the occasional unhappy buyer. It is producing bad experiences every day, whether or not anyone is counting them.
What buyers do after a bad experience
Buyers do not need many chances to leave. Zendesk’s 2025 CX Trends research found that 63% of consumers say they would switch to a competitor after a single bad experience, up 9% on the year before. [3] Zendesk — 2025 CX Trends Report
PwC’s 2025 Customer Experience Survey of more than 5,000 consumers found that 52% have stopped using or buying from a brand because of a bad experience with its products or services, and 29% have stopped because of poor customer experience more broadly. [2] PwC — 2025 Customer Experience Survey
Post-purchase moments carry particular weight. Narvar’s 2025 report, based on a survey of more than 3,400 US online shoppers, found that 76% say they won’t buy again after a poor returns experience, and that 90% check the return policy before they buy. In the same survey, half of consumers said they are less likely to shop with a retailer again after a late delivery. [4] Narvar — 2025 State of Post-Purchase report[5] Narvar — Anxious online shoppers want a happy ending (2025 consumer survey)
These are survey answers about intent, and people do not always do what they say. But they point in the same direction from different researchers, samples and years: a single operational miss is enough to put the next order in doubt.
Why the cost is lifetime, not one order
Here is a simple, hypothetical illustration. It uses round numbers to show the mechanism, not to predict any particular shop’s results.
| Assumption or result | Value |
|---|---|
| Orders a month | 2,000 |
| Average order value | $35 |
| Orders a repeat buyer places in a year | 3 |
| Orders that hit a CX leak | 1 in 9 (about 11%) |
| Buyers who leave after an unfixed leak (assumed) | 50% |
| Buyers lost in a year | About 1,320 |
| Repeat revenue lost with them | About $138,600 |
The order on which the leak happened is a small part of that figure. Most of it is the next two orders the buyer would have placed. Change the assumptions and the total changes, but the shape does not: the lifetime loss is several times the size of the original order.
Replacing those buyers is not free. Harvard Business Review, summarising research across industries, notes that acquiring a new customer costs somewhere between five and 25 times as much as retaining an existing one, and cites Frederick Reichheld of Bain & Company: increasing customer retention by 5% increases profits by 25% to 95%. [7] Harvard Business Review — The Value of Keeping the Right Customers (2014)
On a channel where new buyers often arrive through paid placement, affiliate commission and free samples, a lost repeat buyer is a cost you pay twice: once in the revenue that leaves, and again in what it takes to replace it.
Returns: the leak that costs twice
Returns are both an operational cost and a loyalty moment. The National Retail Federation and Happy Returns projected that US retail returns would total $890 billion in 2024, about 16.9% of annual sales, and found that retailers’ online return rates were 21% higher than their overall return rates. [6] NRF and Happy Returns — 2024 retail returns to total $890 billion
That makes the returns experience a large and recurring test of whether a buyer comes back. A refund processed promptly and explained clearly can keep a customer whose item did not work out. A slow or confusing one turns a product problem into a relationship problem.
Finding leaks while they are cheap
The research points to one practical conclusion: the cost of a leak depends heavily on how long it goes unnoticed. A label caught in the first hour is a two-minute fix. The same miss found a week later is a cancellation, a bad review and a buyer who does not return.
- Count leaks, not just complaints. Track late dispatches, stalled packages, slow refunds and unanswered messages per 100 orders, every week.
- Measure time to notice and time to fix. A falling leak rate matters less than how quickly the remaining leaks are caught.
- Look for the cause behind the pattern. Four “runs small” returns on one product point to a size chart, not four unlucky buyers.
- Close the loop with the buyer. A fast, honest reply after a miss is often what decides whether they order again.
- Review the same few numbers on a fixed day. Leaks are easy to ignore one at a time and hard to ignore on a weekly chart.
Reading the data carefully
Most of the figures here come from consumer surveys run by research firms and software companies. They are useful for direction and scale, but they measure what people say, use different questions and samples, and are not interchangeable. Use them to decide where to look in your own shop, then let your own orders, returns and reviews tell you how large each leak really is.
Sources & further reading
Platform facts are linked above. Examples and operating recommendations are Ruleo’s analysis. Sources checked October 31, 2026.
- [1] Qualtrics XM Institute — $3 trillion is at risk due to bad customer experiences in 2026
- [2] PwC — 2025 Customer Experience Survey
- [3] Zendesk — 2025 CX Trends Report
- [4] Narvar — 2025 State of Post-Purchase report
- [5] Narvar — Anxious online shoppers want a happy ending (2025 consumer survey)
- [6] NRF and Happy Returns — 2024 retail returns to total $890 billion
- [7] Harvard Business Review — The Value of Keeping the Right Customers (2014)