Every resignation feels sudden to the person receiving it. It almost never is. By the time someone hands in their notice, they've usually spent months quietly detaching — taking fewer swings, celebrating fewer wins, drifting to the edge of the team. The frustrating part isn't that the signs weren't there. It's that nobody was looking at the one dataset where they show up earliest.
That dataset is recognition. Treat recognition as an early-warning system for attrition and you stop learning about disengagement from exit interviews — where it's expensive trivia — and start learning about it while there's still time to act. This post covers why recognition activity drops before people do, which specific signals to watch, and exactly what to do in the weeks after the alarm goes off.
Why Resignations Blindside Us
Here's the uncomfortable math: the Work Institute finds that about 3 in 4 voluntary departures are preventable. Not inevitable life changes — preventable. Someone in the company could have done something, and didn't, usually because they didn't know anything was wrong until the calendar invite titled "Quick chat?" appeared.
Surveys don't save you here. Annual engagement surveys arrive once a year, pulse surveys get gamed or ignored, and a disengaging employee is precisely the person least likely to write a candid free-text answer. What you need is a signal that's behavioral (people can't perform enthusiasm in it for long), continuous (it updates weekly, not annually), and ambient (nobody has to fill anything out). Recognition activity is all three.
Why Recognition Drops Before People Do
Recognition is discretionary social investment. Nobody is required to send a shout-out to the teammate who unblocked them — they do it because they're paying attention to the team and emotionally invested in its wins. Which is exactly why it's such a sensitive instrument: discretionary investment is the first thing a disengaging person withdraws, long before their deadlines slip and years before their badge stops working.
The withdrawal happens on both sides of the ledger:
- Giving drops first. Someone who's checked out stops noticing — or stops caring about — other people's good work. A person who gave props weekly for a year and has gone silent for six weeks has changed something about their relationship to the team, whether or not they've admitted it to themselves.
- Receiving drops next. This one is a compound signal: either their contributions are genuinely shrinking, or they've drifted far enough to the periphery that teammates no longer see their work. Both are bad, and both feed the disengagement they reflect. 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 — a receiving drought isn't just a symptom of attrition risk, it's an accelerant.
- Connections narrow last. Before going silent entirely, people often shrink their circle — recognition flowing to and from one or two close colleagues instead of across the team. Social withdrawal in miniature, visible in the data.
None of this requires anyone to confess anything. It's just what disengagement looks like when it leaves footprints. We dug into the mechanics of reading those footprints in detecting quiet quitting with recognition data — this post is about the operational side: turning the pattern into an alarm, and the alarm into action.
The Three Alarms Worth Setting
Alarm 1: The Giving Cliff
Watch for an active giver going quiet. The baseline matters more than the absolute number — someone who never gave recognition going another month without giving is noise; someone who averaged four gives a month dropping to zero for six weeks is signal. Rule of thumb: a 50%+ drop from personal baseline, sustained for four or more weeks, is worth a human's attention.
Alarm 2: The Receiving Drought
Watch for a regular recipient whose inbound recognition dries up. Same baseline logic: the person who reliably collected props for shipping, helping, and unblocking, now going a month or more without a single mention. Before you panic, check the obvious innocent explanations — but don't let "they're probably just heads-down" become the story you tell yourself for a quarter.
Alarm 3: The Shrinking Circle
Watch for someone whose recognition graph is collapsing inward — fewer distinct people giving to them, fewer distinct people they give to. This one often fires before the volume alarms, and it's also where manager perception is least reliable, because the manager may still be having perfectly normal one-on-ones while the peer network quietly dissolves. (Managers miss more than you'd think — we cataloged the patterns in manager blind spots that recognition data reveals.)
One person tripping one alarm is a prompt to look closer. One person tripping two or three, for four-plus weeks, is a retention conversation you should have this week. And if you want the broader checklist beyond recognition data — camera-off meetings, minimum-viable participation, the works — see the warning signs of quiet quitting.
When the Alarm Fires: A Four-Step Playbook
Step 1: Rule Out the Boring Explanations (Days 1–3)
Vacations, parental leave, a brutal crunch project, a role change that moved someone to a new team — all of these dent recognition activity without meaning anything. Check the calendar before you check your assumptions. The goal of this step is to avoid the single worst move available: treating a data blip as an accusation.
Step 2: Have a Real Conversation (Week 1)
Not "the data says you're disengaged" — never that. Recognition data tells you where to look; it should never be the script. The manager books a genuine one-on-one and asks open questions: What's energizing you lately? What's draining you? Is the work still going somewhere you want to go? Gallup attributes about 70% of the variance in team engagement to the manager, which means this conversation — done with actual curiosity — is the single highest-leverage intervention on this page.
Step 3: Reconnect Them to the Team (Weeks 2–4)
If the conversation surfaces drift rather than a hard blocker, engineer visibility. Put them on work that touches other people. Make sure their next real contribution gets recognized publicly and specifically — not a pity shout-out (people smell those instantly), but genuine props for genuine work. Recognition is both the smoke detector and part of the sprinkler system: feeling seen again is itself re-engaging.
Step 4: Fix What the Conversation Surfaced (Weeks 2–8)
Sometimes the drift has a concrete cause — a stalled promotion, a project they hate, a conflict nobody addressed. The early-warning system bought you months of lead time; spend it actually fixing the thing. An alarm you acknowledge and ignore trains everyone, including you, to ignore the next one. For the org-wide version of this discipline — dashboards, cadences, who reviews what — see making recognition data actionable.
The Guardrails: This Is a Flashlight, Not a Camera
Two rules keep this system healthy. First, recognition data is never a performance metric. The moment "props received" affects a review, people will farm it, the data becomes theater, and your early-warning system goes deaf. Second, the signal prompts a conversation, not a conclusion. Plenty of healthy, engaged people have quiet months. You're not surveilling employees; you're noticing when someone might need attention sooner than the annual survey would tell you — the same way a good manager in a physical office notices someone eating lunch alone for a month.
What the Lead Time Is Worth
Why build any of this? Because the alternative is astonishingly expensive. SHRM puts the cost of replacing an employee at 50–60% of their salary; Gallup's range runs from one-half to two times salary. For a 100-person company with a $65,000 average salary and 15% turnover, even the conservative 50% figure works out to $487,500 a year walking out the door. Run your own numbers through our employee turnover cost calculator — the result is usually the moment this stops being an HR curiosity and starts being a budget line.
And the lever works: Deloitte's research ties companies with strong recognition cultures to up to 31% lower voluntary turnover. An early-warning system that helps you save even two or three of those preventable departures a year pays for itself many times over — the retention math is not close.
Setting This Up in Slack
You can't watch a signal you're not generating, so the prerequisite is simple: lightweight, continuous peer recognition where your team already works. Full disclosure — yes, Propsly is ours — but the setup genuinely takes minutes: install it in Slack, and teammates give recognition with the /props command. The free tier covers unlimited users with 200 props per person per month, leaderboards, and a public recognition feed — which means every give becomes a timestamped data point from day one. The Pro plan ($50/month flat for the whole workspace) adds the analytics layer that makes the alarms visible: engagement gaps, participation trends, and the per-person patterns this post is built on.
Whatever tool you use, the principle stands: people tell you they're leaving long before they tell you they're leaving. Recognition data is where they say it first. Build the system that's listening.