Nobody churns at renewal

Renewal season comes round, one of them goes badly, and then there is a slide. The slide says pricing pressure. Or vendor consolidation, or a change of strategic direction at the customer, which is the same slide with different words on it. Everyone in the room nods, because all of those are plausible and none of them are anybody's fault, and the account gets closed out as a pricing problem.

I have sat in a fair few of those rooms now, across acquiring, orchestration and issuing, and I do not think I have ever seen one where the reason on the slide was the reason. The customer had usually made their mind up long before. Somewhere in the previous year, during the actual work, something happened that told them what this relationship was going to be like, and by the time it reached a renewal meeting it was less a decision than an administrative step.

01. The reason on the slide

Price is what customers say because it costs them nothing to say it. It is polite, it is not personal, and it means nobody has to sit through an awkward conversation about the eleven weeks in February and March.

What I have noticed is how differently the same price increase lands depending on where you already stand. A customer who rates you will argue. They will push back hard, tell you it is unacceptable, ask what you are going to give them in return, and eventually you land somewhere that neither side loves. That is a healthy account behaving normally. A customer who has already written you off does not argue. They are polite about it, they say they will take it away and review internally, and a few weeks later someone forwards you an RFP that has clearly been in drafting for a while.

Same email, same number, completely different meaning, and only one of them is really about price.

The sponsor-left explanation is similar. Stakeholders move on all the time. Plenty of relationships survive it fine, because enough people on their side can explain what you are for. When an account dies within a quarter of one person leaving, that person was not a champion so much as the only thing holding the thing upright, and that was true well before they resigned.

02. It starts at integration

The first genuinely honest thing you do with a customer is integrate them. Everything up to that point is sales, and both sides know it. The commercials were agreed by people who will not be in the operational calls, the demo environment was clean, and you picked the reference customers.

Then their engineers meet your engineers. They meet your documentation, your sandbox, your certification process, your ticket queue, your change windows, and whatever the honest state of your escalation path is on a Friday afternoon. That is the first time they experience working with your company rather than being sold to by it, and the gap between those two things is nearly always wider than anyone internally wants to admit.

I have watched a decent commercial relationship get quietly poisoned by a sandbox returning error codes that did not tell you anything useful. Not an outage, not a breach, nothing that would ever appear on a risk register. Just weeks of a customer's engineers guessing, raising tickets, waiting, and forming a view. The view gets said out loud in their internal channel and it is usually about six words long. These people are hard to work with.

That sentence has an extraordinarily long life. It survives account manager changes, product releases, roadmap wins and genuinely good QBRs. New joiners on their side get told it as context before they have ever spoken to you. There is a decent chance that the person who repeats it in the renewal meeting was not even there when it happened.

Which makes it a bit odd that most CS organisations put their strongest people on renewals and their newest on implementations.

03. Quiet accounts

Once you are live, volumes settle, tickets drop away, meetings get shorter and the health score goes green. Everyone reads that as the relationship maturing. Sometimes it is.

The thing I actually watch is whether they are still asking me for things. Customers who intend to keep working with you stay demanding. They want to know what is on the roadmap because they are planning against it. They tell you what they are building before it is announced, because they need to know whether you can carry it. They complain, which takes effort and is therefore a decent sign. Customers who have moved on go quiet and become very easy to manage.

The one that always makes me sit up is finding out about a customer's new product from LinkedIn. Not because it is rude, though it stings a bit. It is that the conversation where they worked out who to build it with clearly happened in a room I was not in. If it has been six months since a customer asked for anything outside the contract, and six months since they told me something that was not already public, I do not care what colour the dashboard is.

04. Fixing it takes longer than you want it to

You cannot repair a bad integration with warmth. Lunches do not do it. Extra QBRs definitely do not do it, and putting a second CSM on the account tends to read as panic rather than seriousness.

The thing that has worked for me is narrower and more uncomfortable than that. You say what went wrong, out loud, to their technical team, with no hedging and no corporate apology language, which everyone recognises and discounts anyway. What broke, why it broke, what has structurally changed so it does not happen again. Serious customers take this well because it is what they would do in your position. The urge to soften it is really just the urge to be liked, and being liked is not the objective in that particular meeting.

Then you do something well that they did not ask for. Small things count for a lot here, because they are evidence rather than promises. One clean piece of work delivered early will shift a technical team's opinion of you further than a year of being steadily adequate.

And then you wait, which is the bit everyone skips. If they had a rough nine months, this is not sorted in six weeks. You are asking a group of engineers to change a view they arrived at through direct experience, and those views move slowly. What usually happens instead is that the first genuinely warm meeting gets read as the account being recovered, everybody relaxes, the touch gets lighter, and six months later you are back where you started with less credibility to spend.

05. Payments makes it worse

Two reasons, I think.

The chain is long and the blame is spread across it. On any given programme you might have a processor, a BIN sponsor, a scheme, a programme manager and a couple of vendors on the customer's side, and when something takes eleven weeks instead of four it is often genuinely nobody's fault. Try explaining that to a customer whose launch date has moved twice. They experience it as your delay, because you are who they signed with and you are who they are on the call with. Walking them through the topology is accurate, and it helps you not at all.

The other one is switching costs, which make everything look safer than it is. Moving an issuing programme is slow, expensive and risky, so account teams assume the customer is stuck and behave accordingly. But high switching costs do not stop customers leaving. They just mean that a customer who is going to spend eighteen months and a serious budget getting away from you has to build a proper internal case for it first, and that gets built quietly. By the time it surfaces it has an exec sponsor, a budget line and a preferred alternative, and you are being told rather than consulted.

So the industries where everyone feels safest are the ones where you get the least warning. I do not think that is widely appreciated.

06. Try this

Take your three biggest customers and, for each one, name the moment the relationship was decided.

Sometimes it is easy. The outage where you were on the call at two in the morning and their head of engineering clocked it. The certification that went through first time when their last provider needed three goes. The week you told them something inconvenient before anyone else would.

If you cannot name it, that does not mean it never happened. It means it happened somewhere you were not, probably in a channel between two engineering teams, and you have spent a couple of years managing a relationship without ever having seen what it is standing on.

07. What I would change

Put good people on implementation. Renewal is the easiest meeting in the cycle when the first six months went well, and close to unwinnable when they did not, so loading the end of the lifecycle with your best CSMs is optimising the wrong quarter.

Watch direction of contact as well as revenue. Even something crude, like the date a customer last told you something that was not yet public, has been a better early signal for me than most health scores, and it is harder to game than a RAG status.

Go to the technical calls even when you have nothing to add. Especially then. Being visibly around while something is hard is most of the job, and it is what customers remember when someone asks them, eighteen months later, whether the incumbent has actually been any good.

And when you do lose one, have another look before the pricing slide goes in. Go back through the timeline and find the month it went wrong. It will be earlier than anyone expects, it will be operational rather than commercial, and it will usually have been perfectly visible at the time to anyone who happened to be looking.