HR’s favorite metric is broken. Here's what to examine instead.
There’s an economic concept called Goodhart’s Law that shows how metrics go wrong. It’s typically given as an adage: “When measurement becomes the target, it ceases to be a good measure.”
The idea is simple enough. If you reward people to sell more cars, they’ll sell more cars, even at a net loss.
The measurement stops tracking the thing you want it to track — revenue realized from selling more cars — and becomes valued for its own sake.
Many HR teams have been operating under a version of Goodhart’s Law for years, without knowing it.
HR has no shortage of metrics. Engagement scores, Employee Net Promoter scores (eNPS), time-to-fill, acceptance rates — the dashboard keeps growing.
But retention has always sat at the top of the heap. Regrettable turnover is easy to measure and makes for a compelling story: losing employees is evidence that something is broken in your organization.
The problem is that the labor market has completely shifted over the past several years, while the measurement stayed the same.
In 2022, employees had an abundance of options. The Great Resignation gave HR something to fear: employees leaving in droves, and organizations reckoning with the question of why.
Regrettable turnover meant something then. But that labor market has vanished. Hiring has cooled. Layoffs are climbing. And employees who might have left two years ago are planted in their cubicles. Your corporate culture didn’t improve. Things on the outside got much worse.
HR is measuring what it’s always measured. But the measurement has become a misguided target rather than a measure. Something Charles Goodhart would recognize.
A high retention rate in 2026 doesn’t tell you people want to be there. What does it tell you? That they’re there.
The ROI of compensation changes
One common reason employees leave is finding a job that pays better. Maybe they didn’t get the pay increase they were expecting and decided to look for other opportunities.
But when there are fewer jobs, a lower pay increase isn’t going to goad employees into action as it otherwise might. Higher paying jobs (or any jobs for that matter) are simply harder to find.
Payscale’s 2026 Compensation Best Practices Report suggests most organizations haven’t updated their measurements to reflect the new market logic. When we asked employers how they measure the ROI of compensation changes, 52% pointed to retention rates.
It was the most common answer in the report, ahead of productivity, engagement, and every other metric. The problem? Retention numbers and even employee engagement scores will look great right up until they don’t.
Employees who feel underpaid and undervalued aren’t leaving right now. They’re staying and recalibrating. Effort begins to slow. Investment in their work, team success, and your organization’s future starts circling the drain.
The kind of engagement that actually changes business outcomes (creative problem solving, genuine ownership of outcomes) doesn’t show up in metrics. More alarmingly, its absence doesn’t either.
By the time the labor market opens back up (which might happen sooner than we think) and employees start testing the job market again, organizations that relied on retention as a proxy for health will find themselves holding the bag.
When job growth finally picks up, turnover will come as an unwelcome ending to anemic engagement.
Return to the actual data
HR's hands aren't tied. Practitioners can still measure the health of their workforce. It just takes more effort and better tools.
Compensation data gives you something retention never could: an understanding of whether the conditions for satisfaction are actually in place.
By the time voluntary turnover becomes a problem again, the damage will already be done. The advantage of knowing where you sit with pay is that it gives you something to act on before the labor market makes a decision for you.
The question is how to get that read with any confidence. Most compensation teams are working with more data than they can synthesize and less clarity than they need.
Compensation data that points you in the right direction
Knowing where you stand is harder than it sounds. Compensation teams have data: market ratios, salary range spread, performance rates, and pay increase averages.
What most of them don’t have is a clear read on whether pay decisions, accumulated across hundreds of offers and merit cycles, are sound.
That’s the gap Payscale Compass was built to close. It’s an intelligence platform organized around four indicators: workforce competitiveness, workforce health, workforce investment, and pay architecture. Each one surfaces a different dimension of the same underlying question: are your compensation dollars going where they should?
Workforce competitiveness measures your compensation percentile ranking against a peer cohort competing for the same talent, broken down by job family, showing where you stand overall. Workforce health adds what retention alone never could: rather than counting who left, it examines tenure patterns for genuine stability.
The other indicators (workforce investment and pay architecture) track whether compensation dollars are flowing toward the roles that matter, and whether your ranges have room to reward the people worth keeping.
What separates Compass from traditional benchmarking isn’t the underlying data. Standard benchmarking tells you what a specific role should be paid. Compass tells you whether your decisions, in the aggregate, are strategically sound or moving in the wrong direction. Right now, that may be the most reliable signal you can track.
Compass doesn’t replace compensation expertise; it makes it defensible. When leadership asks why pay is structured the it is — or whether your organization is competitive where it counts — the answer shouldn’t depend on isolated data pulls or instinct. You need a solution that connects individual pay decisions to organizational outcomes, which is precisely what Compass provides.
Want to see Payscale Compass in action?
