What Your Best People Are Quietly Tolerating

The following is a composite case, assembled from patterns we see often enough that the details are almost interchangeable. The company is real in shape and invented in specifics: a 220-person B2B services firm, profitable, well-liked by its clients, with an engagement score in the top quartile of its benchmark set.

Over eight months it lost two of its four strongest people. Both resignations were described internally as compensation-driven. Both were coded as "market pull" in the leadership dashboard. Neither had much to do with money.

What the exits looked like from the top

The first departure was a senior delivery lead who had run the firm's three largest accounts. She resigned on a Tuesday, gave five weeks, transitioned everything cleanly, and told her exit interviewer that she'd received an offer she couldn't turn down. That was true. It was also not the reason.

The second, seven months later, was a practice lead who had built the firm's fastest-growing service line. Same pattern: professional, generous, unrevealing. Same coding in the dashboard.

Leadership responded the way most leadership teams do. They ran a compensation benchmarking exercise, found they were slightly behind on two roles, and adjusted the bands. The adjustment was defensible and irrelevant.

What the exits actually looked like from below

The finding came six weeks after the second departure, from an unlikely source: a project retrospective on a delayed client engagement. Buried in the notes was a sentence someone had written without emphasis — "we lost three weeks waiting on a scope decision that nobody could confirm they owned."

When the COO pulled the last two years of retros, that sentence appeared, in various phrasings, eleven times. The pattern was consistent and specific: decisions that crossed the line between delivery and sales had no clear owner. Anything in that seam waited for an informal escalation to a founder, which usually took one to three weeks and often arrived with a change in direction.

Both departed leaders had been absorbing that gap personally. Both had built elaborate informal workarounds: standing side conversations with a founder, pre-negotiated fallback positions with clients, buffer weeks quietly baked into every plan. Both were extremely good at it, which is precisely why nobody noticed a problem.

High performers don't complain about broken systems. They route around them — competently, invisibly, and then permanently.

Why the strongest people are the last to complain

This is the mechanism worth understanding, because it inverts the usual assumption that your best people will tell you what's wrong.

• They can fix it themselves. Capable people solve local problems locally. The system never registers a failure, so it never generates a signal.
• Complaining looks like weakness. In most cultures, the person who raises a structural problem is asking to be handed it. High performers already have enough.
• They've priced it as permanent. Once someone concludes that the friction is a fact of the place rather than an open question, the topic moves from "raise it" to "decide whether to stay."
• Their exit is professional. The same conscientiousness that made them valuable makes their resignation uninformative. They will not burn a bridge to give you a diagnosis.

The practical consequence: the absence of complaints from your strongest people is not evidence of health. It's an absence of data.

How to surface it before the resignation

The firm in this case adopted four practices. All of them are cheap, and three of them are conversations.

1. Ask about workarounds, not satisfaction

"Are you happy here?" produces a social answer. "What are you working around?" produces an operational one. The questions that worked:

• What's the workaround you've built that you'd be embarrassed to explain to a new hire?
• Where do you routinely add buffer time, and what are you buffering against?
• What decision do you have to escalate that you think you should just be able to make?
• If you left tomorrow, what would break that only you currently hold together?

That last question is the highest-yield one in the set. It maps your single points of failure and your friction in the same answer.

2. Mine the artifacts you already have

The signal in this case was already written down eleven times. Retrospectives, project post-mortems, slipped-deadline explanations, and QBR notes contain your friction inventory; nobody reads them across time. Once a quarter, someone should read the last quarter's retros looking only for repeated phrases, and write down anything that appears three or more times.

3. Track decision latency in the seams

The firm started logging one number per delayed engagement: days spent waiting on a decision, and who the decision was waiting on. Within a quarter the seam between delivery and sales was undeniable, and it was fixed with a single documented rule about who owns in-flight scope changes under a dollar threshold.

4. Do the exit interview six months early

For a small number of people you'd be badly hurt to lose, run the exit interview while they're staying. Same questions, different framing: what would make you consider leaving, what would you change on your first day as CEO, what have you stopped bringing up.

This only works if something visibly changes afterward. Asked once with no follow-through, it becomes evidence that raising things is pointless — which is worse than not asking.

What changed, and what it cost

The fix was a one-page decision-rights document and a threshold rule: scope changes under a set dollar amount are decided by the delivery lead, full stop, with sales notified rather than consulted. Implementation took a two-hour meeting and a paragraph in the operating handbook.

The cost of not having it was two senior departures, roughly a year of leadership attention, a compensation exercise that addressed nothing, and eleven written signals nobody had read.

That ratio is the point of this piece. The friction that drives out strong people is rarely dramatic and rarely expensive to fix. It's just invisible, because the people carrying it are good enough to keep carrying it right up until the day they stop.