There’s a specific number that shows up in almost no standard board pack, and it’s arguably more predictive of a portfolio company’s health over the following twelve months than most of what does show up: the trend in time-to-resolution for customer-reported issues, tracked at the individual issue level rather than as an average. 

The reason this matters more than it gets credit for is that it’s usually the earliest visible signal of organisational strain, well before that strain shows up in retention or revenue numbers. 

A support or customer success function that’s quietly falling behind on resolution times is often the first sign that headcount hasn’t kept pace with customer base growth, or that a product issue is generating more inbound than the team can absorb, or that internal prioritisation has drifted away from customer-facing problems toward internal projects. 

None of that shows up in NRR yet, because customers who are having a slower-than-usual support experience don’t typically churn immediately, they tolerate it for a while first, sometimes for several quarters, before it eventually shows up as a retention problem that then gets treated as a surprise. 

Most reporting structures don’t track this at the individual issue level, they track average resolution time, which smooths over exactly the kind of tail-end deterioration that matters most, a growing number of issues taking meaningfully longer than they used to, hidden inside an average that hasn’t moved much yet. 

The practical fix is straightforward and cheap to implement: ask for the distribution of resolution times, not just the average, and specifically ask how the tail, the slowest 10 or 20 percent of issues, has moved over the last few quarters. 

A worsening tail with a stable average is one of the more reliable early warning signs available, and it’s almost never in the standard board pack unless someone specifically asks for it.