Do Employee Recognition Programs Work?

Evidence grade: Misapplied. Good evidence exists and is routinely cited for a claim it does not support. This grade describes the published evidence. It is not a prediction about your workforce and not a result RetainScore has produced.

The evidence is real. It is about something else

Recognition is unusual among retention interventions in having proper randomized evidence behind it. That is why the grade here is not insufficient. The problem is the distance between what those experiments measured and the claim they get used to support.

Two studies do most of the work whenever recognition is defended, and neither one measures whether recognition keeps employees.

The flagship experiment measured a single afternoon

Bradler, Dur, Neckermann and Non ran a controlled field experiment on unannounced public recognition and published it in Management Science. More than 300 people were hired for a three-hour data-entry task, and a random subset of work groups received recognition after two hours. Recognition raised performance in the hour that followed, with the interesting twist that most of the gain came from workers who did not receive it.

It is a clean, well-identified result about effort. Nobody in the study had a continuing job, a manager, a career, or the option to resign. Retention and turnover are not measured and not mentioned. A finding about the third hour of a data-entry shift cannot tell you whether a recognition program keeps a nurse or a software engineer next year.

The study that does measure staying power is about volunteers

Restivo and van de Rijt randomly awarded barnstars, Wikipedia's informal peer awards, to 100 of 200 highly active editors and tracked them for 90 days. Productivity rose 60%, and recipients were also more likely to keep contributing rather than drop out.

That is the closest thing to a retention result in the recognition literature, and it comes with two limits that are usually dropped. The participants are unpaid volunteers with no employment relationship, no salary, and no labor market to move within. And the continued-activity finding rests on 19 users who stopped contributing entirely, five of whom had been treated. A real effect on nineteen people is worth knowing and is not a number to scale to a workforce.

What is missing

We looked for a controlled study testing whether a workplace recognition program reduces employee turnover, and did not find one. What exists instead is a large body of cross-sectional survey work reporting that employees who feel recognized say they are less likely to leave. That is a correlation between two things people report about themselves on the same questionnaire, and it cannot separate recognition causing commitment from committed people noticing recognition.

What this does and does not license

It supports recognition as an effort and morale intervention, which is a real thing to want, with unusually good evidence behind it. It supports the specific design finding that recognition given to top performers moved the whole group, which is more interesting than most advice on the subject.

It does not support a retention business case. If you are asked to justify a recognition budget on the grounds that it will reduce turnover, the honest answer is that the evidence for that specific claim is a 90-day study of Wikipedia volunteers, and you should say so rather than cite research shows and hope nobody follows the footnote.

This page replaced one of ours

Until today this URL held a page titled Employee Recognition That Reduces Turnover, which asserts the claim above as settled. It was published and live. Writing this grade meant reading the evidence properly and finding that our own title had gone further than the research does, so the page was replaced rather than quietly left up.

If you run one anyway

Measure it against your own leaving rates rather than against the literature, because the literature is not about your question. Recognition is cheap enough that the case for trying it does not depend on winning this argument, which is worth saying plainly: the grade here is about what the evidence proves, not about whether the intervention is worth doing.

Work through the retention diagnostic to see which factors the research finds predictive, or read the companion grade on whether employee engagement surveys work, which has the same gap between the evidence and the claim made from it.