You launched a recognition program. The kickoff post got a pile of emoji reactions, a few enthusiastic weeks followed, and now someone in a budget meeting is asking the question every program owner eventually faces: is it actually working? "People seem to like it" is not an answer that survives a finance review. The good news is that learning how to measure whether your recognition program is working doesn't require a data science team — it requires four KPIs, a monthly habit of looking at them, and the discipline to distinguish signals that predict the future from ones that describe the past.
This guide covers exactly that: the participation, reach, equity, and retention metrics that define program health, which indicators lead and which lag, and what to do when a number goes the wrong way.
First, Separate Leading Indicators From Lagging Ones
Before touching a dashboard, get one distinction straight, because mixing these up is how programs get killed prematurely.
- Leading indicators are behaviors you can observe now that predict outcomes later: participation rate, recognition frequency, reach, distribution. They move within weeks of a change and tell you whether the program is healthy today.
- Lagging indicators are the outcomes you actually bought the program for: retention, engagement scores, regretted attrition. They move slowly — quarters, not weeks — and confirm (or refute) that the leading behavior paid off.
The classic failure mode: leadership judges a three-month-old program by a lagging indicator ("turnover hasn't dropped yet"), declares it a flop, and cancels it right before the payoff window. The reverse failure exists too — celebrating great participation for two years while never checking whether retention budged. You need both layers: leading indicators to steer week to week, lagging indicators to justify the program's existence once or twice a year.
The Four KPIs That Define Program Health
1. Participation: Are People Actually Giving?
Participation rate is the percentage of employees who gave at least one recognition in a given month. It's the pulse of the program — if nobody's giving, nothing downstream matters.
Measure it monthly and watch the trend, not the single number. Every program spikes at launch; the question is where it settles. A healthy peer program typically stabilizes with a majority of employees giving something each month. If yours settles below a third, the program isn't embedded in the workflow — it's a destination people forget to visit, and forgotten destinations are how most recognition programs quietly die.
One refinement worth the extra column: track manager participation separately. Gallup attributes about 70% of the variance in team engagement to the manager, and managers who never give recognition quietly signal to their teams that the program is optional theater.
2. Reach: Are People Actually Receiving?
Reach is participation's mirror image: the percentage of employees who received at least one recognition in the period. It's arguably the more important number, because the psychological benefit of recognition lands on the receiver — and the risk lands on the people who never receive any.
Gallup and Workhuman found that employees who don't feel adequately recognized are about twice as likely to say they'll quit within a year. Reach tells you exactly how many people are sitting in that risk pool. A program with 80% participation but 45% reach isn't a success — it's a popularity contest with a retention problem attached. Set a target (90%+ reach per quarter is achievable for most teams) and treat every name on the "zero recognitions" list as a follow-up item, not a statistic.
3. Equity: Is Recognition Spread or Concentrated?
Reach tells you who got any recognition; equity tells you how the volume is distributed. The simplest useful measure is a concentration split: what share of all recognition goes to the top 10% of receivers versus everyone else? If a tenth of the company collects half the props, you don't have a recognition culture — you have a fan club.
Slice the same data by team and by role. Concentration often hides at the team level: sales celebrates loudly while the platform team that kept everything running gets silence. Those quiet corners are where disengagement incubates. We've written a full playbook on diagnosing and correcting this in fixing recognition inequality, but the measurement itself is step one — most program owners are genuinely surprised the first time they see their distribution chart.
4. Retention Correlation: Does Any of This Keep People?
This is the lagging indicator that pays for everything else. You're looking for two comparisons:
- Recognized vs. unrecognized leavers. Of the people who resigned in the last 12 months, how many were in your low-recognition group? If departures cluster among the rarely recognized — and they usually do — you have your correlation.
- Before vs. after. Compare voluntary turnover in the year before the program to the year after, ideally by team. Deloitte's research ties strong recognition cultures to up to 31% lower voluntary turnover, which gives you a ceiling to sanity-check your own numbers against.
Correlation isn't causation, and you should say so in the report — it makes the rest of your numbers more credible, not less. But when the pattern repeats quarter after quarter, it's decision-grade evidence. To translate it into money: SHRM puts the cost of replacing an employee at 50-60% of salary, and Gallup's range runs from one-half to two times salary. Even a modest retention improvement covers a recognition budget many times over — run your own figures through our turnover cost calculator, and see the recognition vs. turnover retention math for the full worked model.
Supporting Metrics Worth a Glance
The big four carry the review, but a few secondary signals add texture:
- Frequency: total recognitions per employee per month. Trend matters more than level — a steady slide over three months is your earliest warning of program decay.
- Cross-team share: the percentage of recognition that crosses team boundaries. It doubles as a collaboration-health gauge — silos show up here before they show up anywhere else.
- Specificity: skim a sample of messages quarterly. "Great job!" is noise; "thanks for staying on the incident until the customer was unblocked" is culture. If generic messages dominate, your program needs examples, not more volume.
- Individual trendlines: a previously active person going quiet — giving and receiving less — is one of the strongest early attrition signals you have. That's the premise behind detecting quiet quitting with recognition data.
Build the Review Cadence
Metrics that nobody reviews are just decoration. A workable rhythm:
- Monthly (15 minutes): check participation, reach, and frequency against last month. Note any team whose numbers dropped and ask the manager what changed.
- Quarterly (one hour): pull the equity distribution, the zero-recognition list, and cross-team share. Pick one corrective action — a manager nudge, a values-tag campaign, a shout-out ritual for a quiet team — and actually do it.
- Annually: run the retention correlation, attach dollar figures, and present it wherever budgets get decided.
Two cautions. First, benchmark against yourself before benchmarking against anyone else — your trend is more honest than someone else's average, though external recognition benchmarks are useful for calibrating targets. Second, never publish leaderboard-style metrics as goals. The moment "recognitions given" becomes a quota, people game it, volume rises, meaning collapses, and every number on this page turns to mush. Measure behavior; incentivize sincerity.
Where the Data Comes From
None of this works if recognition happens in scattered DMs and hallway comments, because unrecorded recognition can't be measured. The prerequisite is a single visible channel where recognition flows — which is why Slack-native tools have become the default for this job.
Obligatory disclosure: Propsly is ours, and it was built around exactly this measurement loop. The free tier gives every workspace unlimited users, 200 props per person per month, a public recognition feed, and leaderboards — enough to generate the raw data. The Pro tier ($50/month flat for the whole workspace) adds the analytics layer: participation and reach over time, concentration splits, team breakdowns, and engagement-gap views, so the quarterly review is a dashboard read instead of a spreadsheet archaeology dig. Other tools can produce similar data; the non-negotiable is that some system records every give with a timestamp, a giver, and a receiver.
The Health Check, In One List
If you only remember five questions, make them these:
- What percentage of employees gave recognition this month, and is it trending up or flat?
- What percentage received recognition, and who's on the zero list?
- How concentrated is the volume — top 10% versus everyone else?
- Are low-recognition employees overrepresented among leavers?
- When did you last change something because of these numbers?
That last question is the real test. Measurement isn't the goal — it's the steering wheel. If you want to go deeper on turning these numbers into interventions, making recognition data actionable picks up where this post leaves off, and the rest of our blog covers everything from equity fixes to the retention math in more depth.