Silent Churn: Why Your Best Customers Leave Quietly
Field notes · Diagnostics
Why your best customers leave quietly
The customer who complains is the one you can still save. The customer who leaves quietly is the one you already lost.
Silent churn is the churn nobody puts on a slide. It does not show up as a support ticket that got mishandled, because there was no support ticket. It does not appear in the NPS, because the customer did not fill in the NPS. It shows up as a slow drift in monthly repeat orders, or a gap where a formerly weekly customer stopped appearing – and by the time it is visible in aggregate numbers, the pattern is two quarters deep.
The specific customer who churned quietly rarely writes back to explain. This is what makes the pattern difficult to work on: your best customers are often your quietest ones, and their departure looks like everyone else’s – a person who used to buy, and does not any more.
The signals are there. They are just not where most operations look.
The customer who stops asking questions
A reduction in ticket volume is usually treated as good news. It is often the earliest sign of silent churn.
A customer who has stopped asking is not necessarily happy. They may have concluded that asking does not help. The first time they wrote in, the reply was slow or beside the point. The second time, they got a fix but had to chase for it. By the third time, they have quietly decided that the fastest way to solve their problem is to solve it themselves – or to buy from someone else.
Academic research on contact centres puts numbers to the pattern. A recent study on silent abandonment in text-based contact centres found that between 3 and 70 percent of customers who leave a support queue do so silently – without saying a word to the agent, without leaving a rating, without cancelling. The system usually does not know they left at all.
The metric that captures this at customer level is not “ticket volume.” Ticket volume aggregates every customer into one number and hides the change. The metric that captures this is “return contact rate for customers who wrote in during the last ninety days.” If that rate is dropping while your customer base is stable, you are watching people give up on you in real time.
The same pattern shows up in a smaller signal – average tickets per customer per quarter, tracked as a cohort. When it drifts down across a stable customer base, some of those customers are not becoming easier to serve. Some of them are becoming ex-customers.
Reduction in ticket volume as a warning, not a win
The dashboard cannot tell the two apart from ticket count alone. Fewer tickets can mean the product got better. It can also mean customers have moved on and have not told you.
The way to tell the difference is to look at what happens after resolution. A healthy service operation has customers who come back with a related question weeks later, buy again, refer someone. A less healthy one has customers whose last interaction was a resolved ticket, and no interaction since.
Track “days since last engagement” as a cohort curve, not an average. The average will always look reasonable because your active customers dominate it. The curve tells you what percentage of your best customers have gone silent – and that percentage is your near-term churn risk.
The gap between resolved and satisfied
Every customer service operation closes tickets it has not actually resolved. This is not a moral failure of the agents. It is a structural feature of how ticketing systems reward closure.
Resolution is a moment; satisfaction is an aftermath. The resolved-not-satisfied gap is small in a healthy operation and large in an unhealthy one. It shows up in the second contact – the customer who writes in a week later about the same underlying issue, dressed up as a different symptom. It also shows up in a total absence of second contact from customers who used to be regulars.
The founder’s version of this check is a monthly read of ten closed tickets picked at random, followed by a quick look at whether those ten customers have been active in the last thirty days. If two of them have gone quiet, that is your silent churn cohort. Small numbers, but predictive.
For a fuller version of the same principle – tied to the four-gap framework – see the piece on how to run a customer service audit. The signals live in different places, but the discipline is the same.
The customer who stopped writing in is not happy. They have stopped expecting an answer.
What NPS misses that the tickets show
NPS has a problem that is not talked about often enough in SME operations. It is a self-selection instrument. The customers who fill it in are the ones who already have strong feelings – advocates and complainants. The middle, and the silently-leaving, do not fill it in.
The people you are losing to silent churn are almost never in your NPS sample. This is not a criticism of NPS as a metric – it works for what it measures. It is a criticism of using NPS as your only churn signal, because it is the wrong instrument for detecting the specific customers who are about to leave without saying so.
The tickets show what NPS misses. Read the last five tickets a customer sent, in order. If the tone flattens across those five – from engaged, to transactional, to short, to nothing – that customer is leaving. Not next week. But probably next quarter.
This is not a data science problem. It is a read-the-tickets problem, and it needs someone senior enough to notice the tone shift.
Why the economics make silent churn expensive
The math of retention has been known for decades. A landmark study by Frederick Reichheld of Bain & Company found that increasing customer retention rates by five percent produces a 25 to 95 percent increase in profits, depending on the industry. The Harvard Business Review summary of the underlying research is worth reading in full. The range is wide, but the direction is not: retention beats acquisition on almost every dimension.
Silent churn is expensive not just because you lose the customer. It is expensive because you lose them without knowing why. Classic churn – a cancellation, a complaint, a public review – at least tells you something you can act on. The quiet version tells you nothing. The customer is gone, and the operation has learned nothing from their departure.
For a business that runs on repeat purchase or subscription, this is the compound-interest problem in reverse: small unnoticed departures every month, aggregating into a stable customer base that is quietly shrinking under stable-looking totals.
Building the check into the operation
Silent churn is a signal problem, not an effort problem. Most operations have the data already. It is just being aggregated in a way that hides the pattern.
Three things to build into the monthly review cycle.
One: track return contact rate for the ninety-day cohort. If it drops while your customer base is stable, name it in the review and investigate.
Two: cohort your “days since last engagement” curve, and watch the tail. The tail is where the losses live.
Three: every month, pick ten closed tickets at random and check whether those customers are still active. Two silent departures out of ten is a signal. Five is a fire.
None of these needs a new tool. All of them need someone whose job it is to look for the pattern.
The uncomfortable observation
Most operations celebrate declining ticket volume. Declining ticket volume, on its own, is neither good news nor bad news. It is a question. The answer is either “the product got better” or “the customers stopped bothering.” One of those is a win. The other is an emergency.
An operation that cannot tell the two apart is not measuring the right things. It is measuring the loud ones and missing the quiet ones – which happen to be, disproportionately, the customers you most want to keep.
Where to start on silent churn
The signal that silent churn is happening in your business is rarely a single event. It is a set of small drifts that no one has connected: slightly fewer tickets, slightly older customers going quiet, slightly more resolved-but-not-really tickets, slightly lower repeat orders. Each on its own looks harmless. Together they are the pattern.
If you would rather start from a structured audit than build one from scratch, the CS Readiness Audit Kit is the version we use with SME operators. It works through the review cycle, week by week, and it is designed to surface the quiet signals that a standard dashboard will not.
Run the CS Readiness Audit Kit
A structured review cycle for the signals a dashboard misses – silent churn, resolved-not-satisfied, tone drift. Designed for SME operations that want to catch the pattern before it shows up in revenue.
Explore the Audit Kit →