What Departing Customers Know That Your Pipeline Report Doesn't
Photo: Martin Grube, Public domain, via Wikimedia Commons
Most owner-managed firms can tell you, within a few percentage points, what their win rate was last quarter. Very few can tell you why the three customers who left in the same period went. That asymmetry is worth sitting with: the business invests in understanding the customers it has not yet won, and almost nothing in understanding the ones it has just lost.
The exit interview is the cheapest corrective available. It requires no software, no budget and no campaign. It requires a diary slot, a willingness to hear something unflattering, and the discipline to act on what you hear.
The arithmetic nobody runs
Take a client billed at £1,500 a month. Over a three-year relationship that is £54,000 of revenue, before any referral or upsell. Replacing that revenue is not a like-for-like exercise: the replacement client must be found, qualified, onboarded and taught everything the departing one already knew about how you work. The cost of winning a new logo is almost always higher than the cost of retaining an existing account, and the first year of a new relationship usually carries a thinner margin than the third year of an old one.
Set against that, the cost of an exit interview is roughly 45 minutes of a senior person’s time. There is no other growth activity in a UK SME with that ratio of cost to potential insight.
Why the conversation rarely happens
Three things get in the way, and all three are understandable.
The first is ownership. The person who managed the account often experiences the departure as a personal verdict, and asking them to run the interview invites either defensiveness or a performance of composure that yields nothing useful.
The second is the assumption of knowledge. Ask a managing director why a client left and you will usually hear “price”. Price is the polite answer; it is what a client says when they do not want to explain the real one, which is frequently about responsiveness, consistency, or the quiet sense of being an inconvenience.
The third is the fear of reopening a closed decision. It is a genuine risk — a poorly handled conversation can look like a last-ditch sales pitch — but it is a risk of execution rather than a reason to skip the exercise.
Ask the person with nothing to sell
The interview should be conducted by someone senior who did not manage the relationship: a director, the operations lead, or the founder in a smaller firm. The criteria are credibility and detachment. The client needs to believe the person in front of them can change something; the interviewer needs to have no professional identity invested in the account having been well run.
Where a relationship sits at founder level on both sides, a trusted non-executive or external adviser is often the better choice, precisely because neither party has to manage the other afterwards.
Timing matters. Ask within a fortnight of notice, before the client’s recollection has been smoothed by relief. Thirty minutes is enough. Offer a phone call rather than a meeting — it is easier to decline, and easier to be honest on.
A short script, and the discipline to follow it
Four questions will carry the conversation:
- “Before we get to the decision — what was working well in the first year?”
- “What changed, and when did you first notice it?”
- “If you were advising us on what to fix first, what would you say?”
- “Is there anything we could have done at the point you decided, rather than now?”
The first question is not politeness. It establishes that the relationship had value, and it usually surfaces the standard by which the client judged everything that followed.
Then follow three rules. Do not defend. Do not sell. Do not explain. When the client pauses, wait. The most useful sentence in an exit interview typically arrives after four or five seconds of silence, when the interviewee decides to say the thing they had planned to leave out.
Take notes close to verbatim, and ask permission to quote them anonymously. Under UK GDPR, a short factual note of a business conversation is generally straightforward to hold, provided it is stored with the account record and not treated as free commentary. Do not record the call without consent.
From five conversations to one decision
A single exit interview is an anecdote. Five, read together, are a pattern — and patterns are what change a business.
Log the stated reason in your CRM against a fixed list of categories (price, service, responsiveness, personnel change, product gap, closure, consolidation) so that reasons can be counted rather than remembered. Review the list quarterly.
Then act once, visibly, rather than nodding at everything. If three of five former clients mention that quotes took a week to arrive, that is not purely a service problem; it is a pricing problem, because you are absorbing the cost of a slow process in your margin. If the recurring complaint is that smaller accounts feel deprioritised, the answer may be a minimum engagement size or a lighter service tier rather than an apology.
Give the change an owner and a date. Then write to the departing clients, once, and tell them what changed. A minority will come back. Most will not — but they will remember the follow-up, and the next time someone asks them to recommend a supplier, your name will not be the one quietly removed from the list.
What the exercise will not do
It will not replace regular contact with existing customers, which is where the cheaper signal lives. It will not produce statistically robust findings, and it should not be presented as though it has. And it will cause harm if the feedback is gathered and then shelved: a client who takes the call in good faith and sees no change has been given a reason to repeat their complaint more widely.
Run properly, though, it does something a marketing campaign cannot. It tells you which promise you are breaking, and how often.
