Engagement surveys
Engagement Surveys Are Failing You
An engagement survey produces a score. A manager trying to keep someone needs a name. Here’s what the annual survey can’t see, and what does.
June 6, 2026 · 5 min read
An engagement survey produces a score. A manager trying to keep someone produces a name. No amount of better surveying closes the distance between those two things, because the distance is arithmetic, not craft. Averaging is what the instrument does. The moment you average a team, the one person you most needed to hear from is folded into everyone else and cannot be recovered from the result.
The engagement survey is not useless. But it is widely asked to do a job it was never built for, which is to keep specific people from leaving. Here is what it cannot see, and why that blind spot is the one that costs you. Industry estimates commonly put the cost of replacing one departing person somewhere between $10,000 and $50,000 once recruiting, onboarding and lost productivity are counted, and that bill arrives one person at a time, never as an average.
What the survey is good at
Start fair. A well-run engagement survey tells you something true about the climate of your organization. It surfaces themes. It tracks sentiment over time. It can flag that a department is struggling or that a policy landed badly. For understanding the weather across the company, it has real value.
The problem is that retention is not a weather problem. It is a person problem. And the survey, by design, smooths people into averages.
The four blind spots
It reports the past.
An engagement survey is a snapshot of how people felt when they filled it out, delivered weeks later after the data is processed. By the time the results are in the room, the person who was quietly deciding to leave has often already decided. The survey is a rear-view mirror in a situation that requires a windshield.
It hides the individual in the aggregate.
A department that scores a healthy 7.8 can contain your best engineer sitting at a 3 and on her way out. The average looks fine. The average is the problem. Turnover risk is concentrated in specific people, and the moment you average them together, the signal you most need disappears into the noise of everyone who is content.
It is anonymous, which is the point and the limitation.
Anonymity gets you more honest answers, which matters. But it also means that when the survey reveals someone is struggling, you cannot tell who, and you cannot act. You are left knowing a problem exists somewhere in a group of forty people, with no way to reach the one who needs the conversation.
It is annual, and people are not.
Engagement is not a fixed trait. A person who was committed in January can be drifting by July for reasons that never existed at survey time. A once-a-year measurement assumes a stability that human beings do not have. The thing you are trying to catch moves faster than the instrument that is supposed to catch it.
Why more frequent surveys do not fix it
The common response is to survey more often. Pulse surveys, monthly check-ins, continuous listening platforms. These help with the timing problem, somewhat. They do not touch the deeper one.
A more frequent average is still an average. A pulse survey still tells you that engagement on a team dipped, not which person is at risk or why. And survey fatigue is real: ask people the same questions every two weeks and response quality falls, honest answers thin out, and you end up with more data and less truth.
Frequency is not the missing ingredient. Specificity is.
What actually catches the person who is leaving
The signal that predicts an individual departure is not a team engagement score. It is the gap between how a specific employee sees their situation and how their manager sees it.
When those two views diverge, the employee is living in a different reality than the person managing them. They feel something their manager has not registered. That gap is one of the earliest and most reliable indicators of flight risk, and it is invisible to a survey that only asks one side, anonymously, once a year.
Catching it requires a different approach. You have to look at individuals, not just populations. You have to hear from both the employee and the manager on the same ground, so the gap becomes visible. You have to do it on a cadence that matches how fast people actually change. And the output has to be specific enough to act on: this person, this unmet need, this conversation to have this quarter.
It is worth being clear about why this is worth the trouble. The Work Institute has reported that roughly three in four voluntary departures are preventable, meaning the employer could have done something about them. Preventable is not the same as prevented. It only becomes prevented when someone knows which person to talk to, and what that person actually needs. A score cannot carry that, however carefully it is collected.
That is not an engagement survey with more questions. It is a different instrument for a different job.
The engagement survey can tell you the building feels cold. It cannot tell you which room is on fire. For retention, that distinction is everything.
Anchor looks at one employee at a time, compares their view against their manager’s, and surfaces the gap that surveys average away.