What your churned customers had in common

Budget was reprioritised. Going in a different direction. The person who championed it has left. Not a fit for where we are heading. If you have lost even a handful of customers you have a list like this somewhere, in a CRM field or the back of a board pack, and every line on it is true. Not one of them is a reason.

They are endings, not causes. A budget gets reprioritised when nobody senior is defending the line item. A champion leaves and takes the whole business case with her, because the case was never written down anywhere she was not standing. Those are two different failures wearing the same four words, and only one of them is about your product.

Gartner surveyed 1,503 B2B technology buyers in February and March 2023 and found that 60 percent of those involved in renewing or expanding an as-a-service agreement regret nearly every purchase they make. Hank Barnes, the analyst who wrote it up, adds the part that matters here: the regret "usually has little to do with the provider or the products themselves". If that is the mood your renewal lands in, the exit form is the last place the real reason will show up.

This is not the loss review you run on a deal you never won, which I covered in what to do when you lose, and it is not the renewal conversation itself, which is your first renewal. This is the review you run on customers who bought, stayed a while, and left anyway. They are the most expensive thing you own and the least read.

Line them up, then ignore the reason

Put every customer who has left on one page and give each of them four columns. Not the stated reason. Who they were when they signed: size, sector, how mature their own process was. How they bought: inbound or chased, one champion or a group, fast or slow. What they actually used in the first ninety days. And who inside your company owned them.

The patterns live in those columns. Six of your nine departures came in through one channel. Five never got past the first feature. Four had the same job title as their only advocate. That is the kind of thing a stated reason will never give you, because the customer does not know it, and would not say it if they did.

How do you do this with eleven customers instead of a thousand?

You stop pretending it is statistics. New Zealand has about 617,000 enterprises and MBIE puts 97.2 percent of them in the small business category, so the number of domestic firms large enough to be a serious customer for most B2B products is small, and the number who have already left you is smaller again. You will never have a significant sample. You are not looking for one.

With eleven churned customers you are not running a regression. You are reading a short stack of case notes and asking what keeps turning up.

Three out of eleven is worth acting on if those three look alike in a way the other eight do not. Treat it as a hypothesis rather than a finding, write it down, and check it against the customers you still have. If the pattern is real, some of them will match it, and those are the ones to go and talk to this month.

The useful output is a sentence about who to stop selling to

Most churn reviews end in a list of product gaps, because that is the comfortable answer and it belongs to someone else. The honest output is narrower and harder to say out loud: a kind of customer you keep winning and keep losing.

Writing that sentence down is a revenue decision, not a support one. It changes who gets qualified out, what the first ninety days have to prove, what you charge, and whether the roadmap item you were about to fund is really the thing that is leaking. Fixing one function at a time cannot reach it, because the pattern sits across all of them.

What changes on Monday

Three things, and they are small. Add the two or three disqualifying questions the pattern just handed you to your qualification, so the wrong customer gets a no earlier and cheaper. Give whoever owns onboarding one measurable thing the first ninety days must achieve, taken from whatever your leavers never reached. And write the business case into a document the customer keeps, so it survives the champion who will eventually move on.

Then put a date in the calendar to do it again. A churn review done once is an anecdote. Done every quarter it becomes the only honest record you have of who your product is actually for.

So: what do the last three customers who left you have in common that none of your best ones do?

Sources. Renewal and expansion regret: Gartner, "Gartner Survey Reveals 60% of Technology Buyers Involved in Renewal Decisions Regret Nearly Every Purchase They Make", press release, 14 June 2023, reporting a survey of 1,503 respondents fielded in February and March 2023 among B2B technology buyers at organisations with annual revenue of at least US$50 million or equivalent, across Western Europe, North America and Asia/Pacific. The 60 percent figure and the quoted line from Hank Barnes, Distinguished VP Analyst, that regret "usually has little to do with the provider or the products themselves", both come from that release. Gartner also describes the figure as a 6 percent increase on 2020 without saying whether that is six percentage points or a relative rise, so I have left the comparison out rather than state it either way. No respondent is identified as a New Zealand buyer and no margin of error is published, so I use it as direction on how a renewal feels from the buyer's side, not as a measurement of this market. Enterprise count: Stats NZ, "New Zealand business demography statistics: At February 2025", provisional total of 617,330 enterprises, up 0.5 percent on February 2024, rounded to "about 617,000" above. Small business share: MBIE, "Small business in New Zealand 2025 factsheet", which states that 97.2 percent of New Zealand enterprises are small businesses and is generated from Stats NZ Business Demography data. That factsheet's size bands are zero, 1 to 5, 6 to 19, 20 to 49 and 50 or more employees, and 97.2 percent corresponds to the bands under twenty employees, but the factsheet does not spell that threshold out, so I have quoted its wording rather than define it on its behalf. The step from those two figures to "the number of domestic firms large enough to be a serious customer is small" is my inference, not a claim either source makes. The four-column review, the three-out-of-eleven threshold, the Monday actions and the quarterly cadence come from my own work with New Zealand tech founders rather than from published data. One figure was considered and cut: the widely repeated claim that winning a new customer costs five times what it costs to keep an existing one runs through hundreds of secondary articles with no primary source I could trace, so it is not in this article.

Nick Burns is a fractional CRO for New Zealand B2B tech companies, at home and expanding internationally. He co-founded Emendo and sold it to McKesson, then the 14th-largest company on the Fortune 500, and has since helped 60+ tech companies grow sales.