Skip to content

The signals

Partner decay: the partners going quiet before they go

Almost nobody resigns. A partner just orders a little less often, across a little less of your portfolio, for two or three quarters — and the quarter you finally notice is the quarter it stopped being fixable.

What leaving actually looks like

A partner that ordered every three weeks starts ordering every five. Then every seven. The basket that used to span six product lines comes down to two, and usually the two with the thinnest margin — the ones they cannot get anywhere else. They stop asking about anything new.

Not one of those movements is hidden. Every one of them is sitting in the transaction history your ERP has been writing faithfully for years. The pattern is there months before the revenue moves, which is exactly the window in which a phone call still changes the outcome.

Why the revenue number hides it

Because revenue is a lagging measure, and it looks fine right up until it doesn't. A partner buying less this quarter than last disappears inside a total that went up.

Sort your partner list by revenue — the way every channel report in this industry sorts it — and a decaying partner still sits comfortably near the top. That is precisely where it should not be. The ranking is not wrong about size. It is silent about direction.

The account manager who would make that call has two hundred partners and a spreadsheet exported some time last quarter. They are not failing to notice. Nothing has ever been pointed at the thing they would need to notice.

What XCOUTER Radar reads instead

Cadence and basket composition over time — the shape of the buying, not the size of it. Health is scored on movement, so a partner that has been shrinking for two quarters and a partner that never grew are told apart and treated as the different problems they are.

What reaches an account manager on Monday is a short ranked list of the partners whose behaviour changed, each one carrying the specific movement that flagged it: the orders, the dates, the lines that went quiet. A conclusion you can argue with, not a score with nothing behind it.

The uncomfortable part

We do not publish a retention figure for this yet. The deployments that would produce an honest one are running now, and the number goes here when it exists — including if it comes back smaller than we hoped.

The Scan tells you which signal is bleeding you

Decay is one of five. The Channel Context Scan reads your own transaction history and names the one costing you most — and deployment starts there, not with a platform rollout.

The other signals