The context
A B2B services firm with a broad but heavily skewed client base: a handful of clients generated most of the revenue, while hundreds of smaller clients were followed up little and badly.
The challenge
The concentration on a few names made the company fragile: losing a single large client would have hit hard. At the same time, many clients were slowing their purchases without anyone noticing: there was no signal, no alert, nothing. By the time the decline was obvious, the client was already gone.
What we did
- We segmented clients into A, B and C by value and potential, to stop treating everyone the same and focus attention where it paid off.
- For each client we defined an expected buying behaviour, so a deviation could trigger an early warning before the client disappeared.
- We used AI analysis of purchase patterns to surface slowing clients that would have gone unnoticed by eye.
- For dormant clients we set up a structured re-contact plan, with priorities based on potential, not on the last invoice.
The good part isn't recovering nine clients. It's realising we were losing them as it happened, not six months later.
The head of sales
The result
In four months nine dormant clients who were quietly slipping away were reactivated, and dependence on the top five clients dropped measurably. The company is now more solid and less exposed and, above all, has stopped noticing problems only after the fact.
Cases are anonymised to protect client confidentiality.