How to Reduce Employee Turnover: A People Ops Playbook
Most advice on how to reduce employee turnover starts in the wrong place. It hands you a list of fixes before you know what is actually broken. Free lunches will not help if your best engineers are leaving because their manager has not asked them a real question in six months.
Here is the number that should reframe the problem. Gallup surveyed people who had voluntarily quit in the past year and found that only 17 per cent had ever been asked what it would take for them to stay. Not 17 percent who said no. Seventeen percent who were asked at all.
Turnover reduction is not a strategy problem. It is a diagnosis problem wearing a strategy costume. This playbook walks through the four steps in order: measure what you have, segment it, find the driver, then act on the driver you actually found.
TL;DR
- 42 percent of voluntary leavers say their manager or organisation could have prevented their exit, which means nearly half your turnover is addressable rather than inevitable.
- The company-wide turnover rate is a vanity metric. Segment by regretted vs non-regretted, tenure cohort, manager, and driver before you spend a rupee or a dollar on fixing it.
- Replacement cost scales with role: Gallup estimates roughly 40 percent of salary for frontline staff, 80 percent for technical professionals, and 200 percent for managers and leaders.
- Compensation explains 30 percent of preventable exits. The other 70 percent comes down to how people are managed day to day, which is cheaper to fix and faster to move.
- Turnover is a lagging indicator with a six to twelve-month delay. Track leading indicators such as pulse trend, 1:1 consistency, and stay-interview coverage so you know it is working before the annual number moves.
Why most turnover reduction plans fail
Most turnover reduction plans fail because they treat turnover as one problem with one cause. It is not. It is a bucket that collects several unrelated failures, and the fixes for each one look nothing like the others.
A company losing people in their first 90 days has an onboarding and hiring-fit problem. A company losing three-year veterans has a progression problem. A company losing everyone under one particular manager has a manager problem. Those three companies need three different plans, and a generic list of retention strategies serves none of them.
The second failure is timing. Gallup found that 77 percent of voluntary leavers either quit within three months of starting a job search or never actively searched at all. By the time someone tells you they are unhappy, the decision is usually made.
That is the whole case for building a listening system rather than an exit-interview habit. Exit interviews tell you why the last person left. Continuous listening tells you why the next one is about to.
Step 1: Measure what you actually have
Before you change anything, get an honest baseline. You need one number, then four cuts of that number, then a cost attached to it.
How to calculate your employee turnover rate
Employee turnover rate is the percentage of employees who left during a period, divided by your average headcount for that period, multiplied by 100.
Turnover rate = (Separations during period / Average headcount for period) x 100
Average headcount = (Headcount at start + Headcount at end) / 2
A 220-person company that ends the year with 180 people and lost 34 during it has an average headcount of 200 and a turnover rate of 17 percent. Run it monthly and annually. Monthly catches problems while you can still act on them.
If you are unsure whether you are measuring turnover or attrition, they are not the same thing and the response differs. We break the distinction down in our guide to attrition in HR.
The four cuts that turn a number into a decision
A single company-wide rate tells you almost nothing about what to do next. Cut it four ways and the plan usually writes itself.
| Cut | What are you asking | What does it tell you |
|---|---|---|
| Regretted vs non-regretted | Did we want to keep this person? | Separates the problem you must fix from the turnover that is doing its job. A 17 percent rate that is mostly non-regretted is healthy. |
| Tenure cohort | How long did they last? 0 to 90 days, 3 to 12 months, 1 to 3 years, 3 years plus. | Early exits point to hiring fit and onboarding. Mid-tenure exits point at management. Long-tenure exits point at progression and pay. |
| Manager | Which teams are losing people? | The single fastest diagnostic you own. Turnover concentrated under one or two managers is a coaching problem, not a culture problem. |
| Driver | Why did they leave, in their words? | Tells you which lever in Step 3 to pull. Without this cut, you are guessing. |
Most People Ops teams have the first three cuts sitting in their HRIS already. The fourth is the one nobody has, and it is the one that matters most.
Segmenting by team and tenure is exactly what culture analytics is built for, and it saves you from rebuilding the same pivot table every quarter.
What turnover is really costing you
Replacement cost is not a flat multiple. It scales sharply with role complexity, and averaging it across your whole org will understate the damage from the exits you care about most.
| Role type | Replacement cost | What does that mean at $80k salary |
|---|---|---|
| Frontline/support | Around 40% of salary | Roughly $32,000 per exit |
| Technical professional | Around 80% of salary | Roughly $64,000 per exit |
| Manager or leader | Around 200% of salary | Roughly $160,000 per exit |
Run this against your own leaver list rather than your headcount average. A 200-person company that lost two managers, five engineers, and eight support staff last year is looking at roughly $896,000 on these multiples at an $80,000 average salary. That is the number that gets you a budget line.
Put a number on it in 60 seconds. Our Wellbeing Calculator turns your headcount, average salary, and exits from the last 12 months into an annual cost of turnover, absence, and disengagement. No signup, no email gate.
Step 2: Diagnose the driver before you pick a fix
This is the step that almost every article on how to reduce employee turnover skips. They jump from “turnover is expensive” straight to “offer flexible working.” You cannot fix a driver you have not identified.
1: The six drivers of preventable turnover
Gallup asked 717 people who had voluntarily quit what their employer could have done to keep them. The answers cluster into six drivers, and the distribution is the most useful thing in the whole dataset.
| Driver | Share | What it sounds like in your data |
|---|---|---|
| Compensation and benefits | 30% | Pay compression against new hires, no visible bands, and offers matched only after resignation. |
| Positive manager interactions | 21% | Low 1:1 frequency, no recognition, employees say nobody listens. The largest non-pay driver. |
| Organisational friction | 13% | Broken processes, unclear decisions, “nothing ever gets fixed” in survey comments. |
| Career advancement | 11% | No development plans, promotions feel arbitrary, mid-tenure exits are climbing. |
| Workload and staffing | 9% | Chronic understaffing, burnout signals, and exits clustered in one over-stretched team. |
| Negative manager interactions | 8% | Micromanagement, rudeness, fear. Usually concentrated under specific managers. |
Read that distribution again, because it contradicts the thing most leadership teams believe. Compensation is the single biggest line at 30 percent, but the other 70 percent comes down to how people are managed day to day.
That is good news for a 150-person company. You may not be able to outbid a competitor on salary. You can absolutely outmanage them.
2: Where to find the evidence for each driver
Each driver leaves a different fingerprint. Here is where to look for it, in order of how quickly you can get an answer.
| Evidence source | What it reveals |
|---|---|
| Turnover cut by the manager | Manager-driven exits. Fastest signal you own. Look for teams running 2x the company rate. |
| Pulse survey trend by team | Early warning on friction, workload, and recognition. Drops in team sentiment usually lead exits by one to two quarters. |
| 1:1 frequency and content | Whether the conversations that prevent exits are actually happening, or whether managers are cancelling them. |
| Compensation band audit | Pay compression. Compare tenured staff against your last three external offers for the same level. |
| Stay interviews | The only source that tells you what would keep someone before they decide to go. |
| Exit interviews | Useful for pattern confirmation, useless for prevention. Treat as a lagging check, not a diagnostic. |
If you are running annual engagement surveys only, you will find the driver about nine months after it stopped mattering. Continuous pulse surveys close that gap, and engagement and retention analytics connect the sentiment trend to actual flight risk by team.
3: The stay interview nobody runs
Return to the statistic from the opening. Among people who quit and had any conversation with a manager in their final three months, only 17 percent were asked what it would take for them to stay. Only 28 percent were asked about their job satisfaction.
Meanwhile, 45 percent of leavers report that no manager or leader proactively discussed their satisfaction, performance, or future with them at all in those three months. The conversation that prevents the exit is simply not happening.
A stay interview is not complicated. It is a 30-minute conversation, run twice a year with every person you would be sorry to lose, built around four questions.
-
What makes you want to stay here right now?
-
What would make you consider leaving if a recruiter called this month?
-
What part of your job frustrates you most, and what have we not fixed?
-
What do you want to be doing in two years, and does that path exist here?
Log the answers. Aggregate them across the team. You now own the fourth cut from Step 1, the one your HRIS does not have.
Step 3: Act on the driver you found
Match the intervention to the driver. Doing all six at once dilutes every one of them and makes it impossible to tell what worked.
1: Fixing pay and progression
Start with a compression audit, not a market benchmark. The most common failure is not that you underpay the market. It is that you underpay tenured people relative to what you just offered a new hire at the same level.
Publish your bands and the criteria for moving between them. Ambiguity about pay reads as unfairness even when the numbers are fine. Then commit to reviewing pay on a calendar, not on a resignation.
Counter-offering a resigning employee is the most expensive way to discover you were underpaying them. It also teaches the rest of the team that resigning is the fastest route to a raise.
Modelling raise scenarios against a fixed budget is what compensation intelligence handles, so you can see the retention impact of a band adjustment before you commit the spend.
2: Fixing the manager relationship
This is the highest-leverage lever you have, and the cheapest. Gallup found that one meaningful conversation per week makes an employee four times as likely to be highly engaged, whether they work onsite, hybrid, or fully remote. The conversation can run 15 to 30 minutes.
Set a floor, not a mandate. Every direct report gets a 1:1 at least fortnightly, it does not get cancelled, and it is not a status update. It covers goals, recognition for recent work, blockers, and growth.
Then measure whether it is happening. Managers who cancel 1:1s are almost always the same managers whose teams show up in your turnover cut, and that correlation shows up long before the resignations do.
Recognition sits inside this lever, not next to it. Twenty-one percent of preventable exits trace to a lack of positive interaction with a manager, and structured recognition makes that visible rather than leaving it to whoever remembers. AI-prepared 1:1 talking points remove the “I did not know what to ask” excuse entirely.
3: Fixing organisational friction and workload
Friction and workload account for 22 percent of preventable turnover between them, and they share a root cause. Someone raised the problem, nothing happened, and they concluded that nothing ever will.
The fix is a visible loop. Collect the friction, publish what you are fixing, publish what you are not fixing and why, then report back on the ones you closed.
A survey with no visible response does more damage than no survey at all. It converts a solvable frustration into evidence that the company does not listen.
See flight risk before the resignation letter. Pulsewise connects pulse sentiment, 1:1 cadence, recognition, and goals into one signal, so you can see which teams are drifting while there is still time to act. Explore engagement and retention.
Step 4: Prove it is working before the metric moves
Here is the trap that kills retention programmes. Turnover is a lagging indicator. If you fix your driver in January, the annual turnover number will not reflect it until late in the year, and your CEO will ask whether it is working in March.
So you need leading indicators. These move in weeks, not quarters, and they are what you take to the board while the lagging number catches up.
| Leading indicator | Cadence | What good movement looks like |
|---|---|---|
| Team pulse trend | Weekly or fortnightly | Sentiment in your worst-performing teams stops falling, then turns. Usually visible in 4 to 6 weeks. |
| 1:1 completion rate | Monthly | Above 90% of scheduled 1:1s are actually held, with no manager below 75%. |
| Stay interview coverage | Quarterly | Every regretted-loss-risk employee has had one in the last six months. |
| Recognition frequency | Monthly | Every team member receives recognition at least monthly. Watch for the teams at zero. |
| Internal mobility rate | Quarterly | Rising share of roles filled internally. Directly counters the career-advancement driver. |
| Regretted turnover rate | Quarterly | The lagging number, but cut to the segment you are actually treating. |
Report the leading indicators monthly and the lagging ones quarterly. It buys you the runway to let the intervention work.
What a good turnover rate looks like
Ten percent or lower is the number most benchmarks land on, and across US industries, the average annual turnover rate sits closer to 18 percent. Both figures are close to useless on their own.
Your target depends on your industry, your growth stage, and your role mix. A hospitality business at 40 percent may be performing well. A 200-person SaaS company at 25 percent has a serious problem.
The more important point is that the right target is never zero. Some turnover is functional. When a poor-fit hire leaves, the team gets stronger, and a seat opens for someone better.
So set your target on regretted turnover, not total turnover. A rate of 17 percent that is mostly non-regretted is a healthier organisation than a rate of 9 percent where every leaver was someone you fought to keep.
A 90-day starting sequence
If you are staring at a high turnover number and a small team, do not attempt all of this at once. Run it in this order.
| Window | Focus | What you do |
|---|---|---|
| Weeks 1 to 2 | Baseline | Pull 12 months of leavers. Cut by regretted status, tenure cohort, and manager. Cost it using the role multiples above. |
| Weeks 3 to 4 | Listen | Launch a short pulse survey to the two teams with the worst turnover. Measure only. Resist the urge to fix anything yet. |
| Weeks 5 to 8 | Diagnose | Run stay interviews with 10 to 15 people you would be sorry to lose. Compare what they say against your pulse data and your lever cuts. |
| Weeks 9 to 12 | Act on one driver | Pick the single largest driver you found. Ship one change. Set the leading indicators from Step 4 and report on them monthly. |
One driver. One change. Measured properly. That beats a twelve-point retention strategy that nobody can attribute anything to.
If engagement turns out to be your underlying driver, the 15 strategies in our engagement guide pick up where this playbook ends.
The bottom line
You will not reduce employee turnover by adding retention initiatives. You will reduce it by finding out which of the six drivers is actually costing you people, then fixing that one properly while you measure whether it worked.
The uncomfortable part is that the biggest lever is not a budget line. It is whether your managers are having a real conversation with each person on their team regularly, before anyone starts looking. Forty-two percent of the people who walk out of your door believe you could have stopped them.
Start with your leaver data this week. Cut it by the manager. You will probably know where the problem lives before you finish the spreadsheet.
Stop finding out in the exit interview
Pulsewise brings pulse surveys, feedback, recognition, goals, and 1:1s into one platform, so you can see flight risk building instead of reading about it on the way out. It replaces five separate HR tools, and it is free forever for the first 100 teams, up to 100 users, with no credit card required. Claim your free spot or schedule a 20-minute demo
FAQs
What is a good employee turnover rate?
Most benchmarks treat 10 percent or lower as strong, while the US average across industries sits nearer 18 percent. Context matters more than the benchmark. Judge yourself on regretted turnover within your own industry and role mix, because a low total rate can still hide the loss of the people you most wanted to keep.
How do you calculate employee turnover rate?
Divide the number of employees who left during a period by your average headcount for that period, then multiply by 100. Average headcount is your starting headcount plus your ending headcount, divided by two. Calculate it monthly as well as annually so you can act on problems while they are still moving.
What are the main causes of employee turnover?
Gallup research into voluntary leavers points to six drivers: compensation and benefits (30 percent), too few positive manager interactions (21 percent), organisational friction (13 percent), lack of career advancement (11 percent), workload and staffing problems (9 percent), and negative manager interactions (8 percent). Roughly 70 percent of preventable exits trace back to day-to-day management rather than pay.
How can managers reduce employee turnover?
By having the conversation before the resignation. Gallup found 45 percent of leavers had no proactive discussion with a manager about their satisfaction or future in their final three months. One meaningful 1:1 per week makes employees four times as likely to be highly engaged, and it can take as little as 15 minutes.
How long does it take to reduce employee turnover?
Expect six to twelve months before the annual turnover number moves, because turnover is a lagging indicator. Leading indicators shift much faster. Team pulse sentiment typically responds within four to six weeks, and 1:1 completion rates move immediately, so track those to prove the intervention is working.
Is some employee turnover good?
Yes. Functional turnover, where a poor-fit or low-performing employee leaves, strengthens the team and opens a seat for a better hire. Targeting zero turnover is a mistake. Set your goal on regretted turnover, meaning the departures of people you actively wanted to keep, and let the rest be.
