Your employee retention rate is the percentage of the people who worked for you at the start of a period who were still there at the end of it. That’s the whole definition. The arithmetic takes about ninety seconds.
The reason it’s worth ninety seconds is that on a small hourly team (a prep crew, a packing bench, a couple of drivers), turnover is invisible until it’s expensive. You replace one person, then another, then you’re training someone in the third week of a busy month and wondering why the customer care standards you set keep slipping. The rate is the number that makes the pattern visible before it gets there.
One note before the arithmetic, because the phrase is ambiguous: this is the employee retention rate, the staffing metric. If you came looking for the customer side of the word, that lives in customer retention examples you can actually use instead.
This post covers the formula, the window to measure it over, what counts as a good number in food and delivery work, and which fixes actually move it.
The Bottom Line
- Employee retention rate = (employees at the end who were also there at the start ÷ employees at the start) × 100. New hires are excluded from both halves.
- Retention and turnover don’t sum neatly to 100%, because turnover counts every departure while retention counts only the original headcount that stayed.
- Measure it over 12 months for the headline figure and over 90 days for new hires. The 90-day number is the one that tells you whether your hiring and onboarding are broken.
- Accommodation and food services has the highest quit rate of any industry the BLS tracks, so benchmark against your own sector rather than the national average.
- Replacing one employee costs roughly a third of their annual salary, about $15,000 for a median-wage worker, per Work Institute’s research.
Lower your delivery costs by 23%
How we reduce costs:
- No delivery vehicle expenses
- Optimized local routes
- Pay-per-delivery model
- Average 23% delivery cost reduction
The employee retention rate formula
There is one formula and one common way to get it wrong.
Employee retention rate = (E ÷ S) × 100
Where S is the number of employees on the payroll at the start of the period, and E is how many of those specific people are still there at the end.
The mistake is putting your ending headcount in the numerator. If you started with 10, lost 3 and hired 4, your ending headcount is 11, and if you divide 11 by 10 you get a retention rate of 110%, which is nonsense. New hires never count. You’re measuring whether the original group stayed, not whether the payroll grew.
A worked example. A catering kitchen starts the year with 14 people: 6 in prep, 4 on the line, 2 drivers, 2 in the office. Over the year, 2 prep staff and 1 driver leave. They hire 5 new people.
- Employees at the start (S) = 14
- Original employees still there at the end (E) = 11
- Retention rate = (11 ÷ 14) × 100 = 78.6%
Headcount ended at 16, which looks like growth. Retention was 78.6%, which means roughly one in five of the people who knew how the kitchen ran walked out of it. Both things are true at once, and only one of them shows up in a headcount report.
Which window to measure over
Run it on two clocks.
- Rolling 12 months for the headline number. Anything shorter and seasonal swings will dominate. A caterer measuring October to December will conclude the business is in crisis.
- 90 days, for new hires only. Calculate what share of the people you hired in a quarter are still there three months later. On hourly teams this is usually where the damage is, and it’s a different problem with different fixes: a bad 12-month rate is a retention problem, a bad 90-day rate is a hiring and onboarding problem.
If you only ever calculate one, calculate the 90-day one. It moves faster and it responds to changes you can actually make this month.
Retention rate vs. turnover rate
People treat these as the same number flipped, and they’re close but not identical. The difference matters when you’re comparing yourself to a benchmark.
Turnover rate counts departures over a period, divided by the average headcount. It includes people who were hired and left inside the same period. Retention rate counts stayers from the opening roster, and ignores anyone hired mid-period entirely.
So a business that churns through six seasonal hires in a quarter while its core staff stays put can post a high turnover rate and a high retention rate simultaneously. Both are accurate. They’re answering different questions: turnover asks how much traffic there was through the door, retention asks whether the people you had stayed.
Use retention when you care about continuity and institutional knowledge: who knows the routes, who knows which customer needs the back entrance. Use turnover when you’re costing out recruiting and training load. Most benchmark data published by the BLS and by HR vendors is turnover data, so check which one a figure is before you compare yourself against it.
What counts as a good employee retention rate
The honest answer is that it depends almost entirely on your sector, and that generic advice on this point is worse than useless.
A widely repeated rule of thumb puts a “good” retention rate at 90% or better, a 10% annual turnover. That target is realistic for salaried professional roles. For an hourly kitchen or packing team it is not, and holding yourself to it just teaches you to ignore the number.
| Sector | Typical annual turnover | Implied retention | Notes |
|---|---|---|---|
| Accommodation and food services | Highest of any sector the BLS tracks; commonly cited near 70–75% | ~25–30% | Quit rates run several times the all-industry figure |
| Retail and wholesale | Around 25% voluntary | ~75% | Seasonal peaks distort short windows badly |
| Transportation and warehousing | Below the food-service figure, above professional services | ~80%+ | Driver roles churn faster than warehouse roles |
| All industries, US average | Roughly 40% total, ~13.5% voluntary | ~60% total | Voluntary-only figures look far better than total |
| Professional and salaried roles | 10–15% | 85–90% | Source of the “90% is good” rule of thumb |
Figures are approximate and drawn from BLS JOLTS reporting and industry summaries; definitions vary between sources, so treat these as ranges to orient against rather than precise targets.
The useful comparison isn’t against the table. It’s against yourself last year, and against the same role in your own business. If prep retention is 80% and driver retention is 45%, you don’t have a retention problem. You have a driver problem, and averaging them together hides it. Always calculate by role once you have more than about eight people.
Why hourly delivery and food work sits at the bottom of the benchmarks
Worth understanding, because the causes point at the fixes.
Accommodation and food services consistently posts the highest quit rate of any industry in the Bureau of Labor Statistics’ Job Openings and Labor Turnover Survey. Some of that is structural and you can’t fix it: the work is physical, the shifts are early or late, and a large share of the workforce is in the job as a stage rather than a career.
But a large share is fixable, and it clusters in four places.
The schedule is unpredictable. Hourly workers organize childcare, second jobs and transport around a rota. A rota that lands on Saturday night for the week starting Monday is a recurring, unforced reason to leave.
The first two weeks are unstructured. A new packer who spends day one shadowing whoever is least busy learns the job slowly and concludes nobody was expecting them.
There’s no visible next step. Not everyone wants a career path, but everyone wants to know whether one exists.
The direct supervisor is untrained. On small teams the supervisor is often the best operator rather than the best manager, promoted with no guidance at all.
What a low employee retention rate costs
The cost is not the job ad. Work Institute’s research puts the cost of replacing a single employee at roughly 33% of their annual salary, about $15,000 for a worker at the US median wage, as reported by HR Dive.
For an hourly role, the money breaks down roughly like this:
- Recruiting and screening time, most of it yours or a supervisor’s rather than a recruiter’s.
- The empty seat: overtime for whoever covers, or output you simply don’t produce.
- The ramp. A packer is slow for two weeks and error-prone for four. A driver who doesn’t know the route runs long and arrives late.
- The errors during ramp, which land on customers. This is the cost nobody puts in the spreadsheet, and on a delivery-driven business it’s often the biggest one, because a wrong or late drop can cost a whole account rather than one order.
That last point is why retention belongs next to your customer metrics rather than filed under HR. The same mechanics apply on both sides of the business. The customer retention tactics worth copying exist because keeping an existing relationship is cheaper than building a new one, and that’s exactly as true of the person driving your van as it is of the café they deliver to.
Fixes that actually move the number
In rough order of how much they move the rate per dollar spent on a small hourly team.
Post the schedule two weeks out and stop changing it. This is the highest-return change available to most operations and it costs nothing but discipline. Predictability is worth real money to hourly staff, and unpredictability is one of the most common reasons they quit for a job paying the same.
Structure the first 30 days. Not a handbook. A plan. Who trains them on what, on which day, and who they eat lunch with in week one. A named buddy for the first month is unreasonably effective for how little it costs.
Fix the pay bands you already know are wrong. You usually know. If a new hire negotiates their way above someone who’s been there two years, you’ve created a departure and set its date.
Train the supervisor, not just the crew. The relationship with the immediate manager is the most consistently cited driver of whether people stay. One day of actual management training for whoever runs the floor outperforms most perks.
Recognize specific work, quickly, in a way that has weight. Generic praise doesn’t register. Naming the specific thing does, like the driver who called ahead when the loading bay was blocked, and small immediate bonuses land harder than an annual review because the connection to the behavior is obvious. If you want to formalize it without building a program, a spot bonus is the usual mechanism: a small, discretionary reward issued within days of the thing it’s for.
Ask the people who stay, not just the ones who leave. Exit interviews collect data from people who’ve already gone and have no incentive to be candid. A ten-minute stay conversation at 60 days (what’s working, what’s annoying, what nearly made you not come back) gets you information while you can still act on it.
Give the job a ceiling higher than itself. A lead-hand title with a small differential, first pick of shifts, or training on a second station. People stay for a visible next rung far more than for a vague promise of growth.
What generally doesn’t move it: pizza, ping-pong, an engagement survey with no follow-up, and an annual bonus far enough away that nobody connects it to today’s shift.
How often to recalculate
Quarterly for the 12-month rolling number, monthly for the 90-day new-hire number. Log it in the same place you log everything else and write one line next to it about what changed.
The value isn’t the figure in isolation; 78.6% means very little on its own. It’s the direction, and whether the direction responds when you change something. If you move the schedule to two weeks out in April and the 90-day retention rate on hires from May onward improves, you’ve learned something you can spend money on with confidence.
Frequently asked questions
How do you calculate the employee retention rate?
Divide the number of original employees still employed at the end of the period by the headcount at the start, then multiply by 100. Exclude anyone hired during the period from both figures.
What is a good employee retention rate?
Around 90% is the common rule of thumb, but it reflects salaried professional roles. Hourly food service and delivery operations run far lower, and the meaningful comparison is against your own previous periods and your own roles rather than a national average.
Is retention rate just the opposite of turnover rate?
Not exactly. Turnover counts all departures against average headcount, including people hired and lost within the same period. Retention counts only how many of the opening roster stayed. The two can move independently.
Why is my retention rate over 100%?
You’ve used ending headcount instead of surviving original employees in the numerator. New hires don’t count toward retention.
What’s the fastest way to improve employee retention on an hourly team?
Publish the schedule further ahead and stop changing it, then structure the first 30 days for new hires. Both are cheap, and both target the periods where hourly staff most often decide to leave.
Start with one number
Pull your payroll for 12 months ago, count how many of those names are still on this month’s payroll, and divide. Then do the same for everyone you hired in the last quarter.
Two numbers, ten minutes. If the second one is much worse than the first, and on an hourly team it usually is, your problem is the first month of employment, not the job itself. That’s a far better problem to have, because it’s the one you can fix by Friday.
