Engagement is not a perk problem

Engagement is not about perks, surveys, or buzzwords.

When turnover starts hurting, most companies reach for the same three things: an engagement survey, a perks budget, and a platform to manage both. The fruit delivery arrives. Someone organizes a quarterly social. And turnover stays exactly where it was, because none of it touched the reason people are leaving.

People leave because they don’t know what’s expected of them, because their manager has never been taught to manage, because they can’t see a path, or because something happened and nobody dealt with it.

The evidence for the survey, the perks and the platform is weaker than the spending suggests. The evidence for something much cheaper is strong, and it’s the thing almost nobody does.

What companies buy, and what actually works

The single largest factor is the manager. Gallup attributes at least 70% of the variance in team engagement to the person a team reports to.

Not the mission statement. Not the benefits package. Not whether there’s cold brew. The manager accounts for most of the difference between a team that’s engaged and one that isn’t, which means most engagement budgets are aimed at the other 30%.

What the manager is actually doing comes down to two things, both measurable and both unglamorous.

Clarity. Only 49% of employees strongly agree they know what’s expected of them at work, as of the first half of 2026 — still well below the 61% peak in 2015 (Gallup). That’s a majority of people going to work slightly unsure what a good job looks like. It costs nothing to fix and it is skipped constantly, because managers assume it’s obvious and employees don’t want to ask twice.

Feedback cadence. Gallup found 80% of employees who had received meaningful feedback in the previous week were fully engaged. Not an annual review. The previous week.

Neither requires a budget. Both require a manager who knows how to do them.

Managers are struggling, and it’s getting worse

Gallup’s State of the Global Workplace shows manager engagement falling from 31% in 2022 to 22% in 2025 — a nine-point drop, with five of those points arriving between 2024 and 2025 alone. Over the same period, individual contributor engagement barely moved: 20% to 19%.

Managers are absorbing almost the entire decline. That makes the standard response — asking managers to improve engagement — a fairly optimistic request of people who are themselves disengaged and untrained.

The same research has the encouraging half: in Gallup’s best-practice organizations, 79% of managers were engaged, nearly four times the global average. This isn’t a fixed cost of doing business. It’s a capability gap, and it’s why manager training is where most of our retention work ends up.

The gap between having a policy and running it

This is the finding closest to companies your size, from research weighted toward them: in Brightmine’s Company Culture 2025 survey, 59% of respondents had fewer than 250 employees.

83% of companies have a documented set of company values. Only 57% train hiring managers to consider those values when hiring. 47% require no training on psychological safety, and at 15% of companies, employees generally don’t feel safe raising constructive criticism.

Read those together. The values exist. The training doesn’t. And in roughly one company in seven, people have concluded it isn’t safe to say so.

An engagement survey run in that environment measures how willing people are to answer honestly, which is not the same thing as engagement.

What the business case actually is

Gallup’s Q12 meta-analysis — 183,806 business units across 53 industries and 90 countries — compares business units in the top quartile for engagement against the bottom quartile:

• 23% higher profitability
• 18% higher sales productivity
• 78% lower absenteeism
• 51% lower turnover in low-turnover organizations, 21% lower in high-turnover ones

One caveat that matters. This compares groups; it doesn’t prove engagement caused the difference. Profitable companies can afford things that engage people, so the relationship may run both ways. Gallup’s own methodology notes say as much, and anyone quoting these figures as a guaranteed return hasn’t read them.

It’s still the best evidence available. It’s evidence of association, not a lever with a known output.

How we work: diagnosis before any survey

We start with what you already have. The pattern is usually visible before anyone is asked anything new.

Turnover by manager. Not company-wide — by team. Turnover is almost never evenly distributed, and a company-wide rate averages a problem team in with five healthy ones.

Turnover by tenure. People leaving inside twelve months is a hiring and onboarding problem. People leaving at three years is a progression problem. Different causes, different fixes, and the company-wide rate hides both.

Exit interview themes, read in aggregate rather than one at a time. Individually they’re polite. In a stack of twelve they’re specific.

Stay interviews. Short conversations with people who haven’t resigned. Faster than a survey, harder to ignore, and the only diagnostic on this list that can change the outcome, because the person is still there.

Then we work on what the diagnosis points to:

The manager layer. Since that’s where most of the variance is, it’s where most of the work is.

Explicit expectations, which cost nothing and are the most commonly skipped step.

Whatever the pattern names. If people are leaving inside twelve months, the fix sits in recruiting, hiring and onboarding. If the pattern is progression or pay, it sits in performance and compensation.

Then, if you still want a survey

Run one only if you intend to act on it.

A survey that produces no visible change teaches people that telling you things doesn’t work — which is worse than never asking. And given that 15% figure above, there’s a real chance a first survey measures caution rather than sentiment.

How long any of this takes

Manager behavior and clarity of expectations tend to move engagement measures within a quarter.

Turnover lags by two to three quarters, because people who have already decided to leave still leave. Your numbers can get worse before they get better, and that is not evidence the work failed.

Anyone promising faster is guessing.

Where we stop

We don’t sell an engagement platform and we don’t take a commission on one. If a tool is the right answer we’ll help you choose it, but most companies under 150 people don’t have a measurement problem. They have a manager capability problem and a clarity problem, and software solves neither.

We’re also not going to promise a number. Anyone who tells you they’ll cut turnover by a specific percentage is guessing, and you should ask them where the figure comes from.