Churn rate

Logo churn counts clients. Revenue churn counts money. In an agency, where account sizes vary by an order of magnitude, those two numbers routinely tell opposite stories about the same quarter.

An agency that lost three small accounts and kept its largest had a good quarter by revenue and a bad one by count. One that lost a single major client and kept everything else had the reverse. Reported alone, either number misleads, and which one gets reported tends to depend on which looks better.

They also answer different questions. Logo churn is about the offer and the delivery model: are clients of this type, at this size, well served. Revenue churn is about the business: can it fund itself next quarter.

The useful habit is to report both every time, in the same place, and to name the accounts behind them. Churn expressed as a percentage is a statistic; churn expressed as three names is something an agency can act on.

Agency churn divides into circumstance and decay, and from the outside they produce identical emails.

Circumstance is the client business changing: budget cut, a marketing director arriving with their own agency, a decision to hire in-house, an acquisition, a pivot. None of that is a comment on the work, and almost none of it is preventable by doing the work better.

Decay is the relationship eroding while the work continues: the agency becoming invisible to the budget holder, a quarter of unaddressed frustration, a contact leaving and the replacement never being properly met, value never being restated.

Sorting each loss into one of the two is the most useful post-mortem an agency can do, and it has to be honest or it is worthless. Filing every loss under circumstance is comfortable and produces no learning. Filing every loss under decay is demoralising and equally false.

The proportion tells you what to fix. Mostly circumstance suggests a targeting problem: the agency is winning clients whose situations do not last. Mostly decay suggests an account management problem, and that one has known solutions.

Churn is almost never a surprise to the people close to the account, and almost always a surprise to the people looking at reports. The gap is because the early signals are operational and the reported ones are performance.

What moves first: approvals sitting unanswered longer than they used to. Meetings rescheduled, then shortened. A new name copied on emails without introduction. Questions that were previously asked by phone arriving in writing. Enthusiasm going out of the responses.

None of those appears in a results dashboard, and several of them happen while performance is fine. An account hitting every target can be lost, because the numbers were never the thing that was decaying.

Which is the argument for a weekly review of account status that includes these signals explicitly, with a named owner and a next action for anything that is not fine. Weeks of warning is normal. Using them is not.

Negative revenue churn means expansion from remaining clients exceeded everything lost, so the existing base grew without a single new client. It is the strongest structural position a recurring-revenue business can be in.

It is harder in most industries than it is in agencies, and the reason is specific: there is nearly always more work to do for a client who trusts you. A software product has a ceiling set by seats and tiers. An agency relationship has a ceiling set by the client marketing budget, which is usually much further away than the current fee.

What stops agencies reaching it is not opportunity, it is that nobody is responsible for finding it. Expansion sits between sales and delivery, and the account team is measured on delivery rather than on growth.

Measuring it is what starts to fix it. An agency that reports net revenue retention alongside churn is asking, every month, whether the base grew, and that question tends to produce the conversations that make it grow.