How to measure guard turnover in your own company
No agency publishes a guard turnover rate. Calculate your own, split it by tenure to find the ninety-day cliff, and cost one departure from your own ledger.
Everyone in this business quotes the same number: guard turnover runs 100 percent a year, sometimes worse. Before you spend a dollar fixing it, find out where that figure comes from and what yours actually is.
You cannot cut turnover you have not measured. Calculate your own annualized rate, then split it by tenure. Losses concentrated in the first ninety days are a hiring problem rather than a pay problem. Cost one departure from your own ledger and you have the budget argument.
The 100 percent figure is a trade rule of thumb, not a measured statistic. The federal turnover survey, JOLTS, publishes hires and separations by industry and by state, not by occupation, so there is no official rate for security guards to look up. Yours is the only one that matters, and you can calculate it this afternoon.
How do you calculate your actual turnover rate?
Separations divided by average headcount, over twelve months, times one hundred. That is the whole formula.
Average headcount is the part people get wrong. Take your headcount at the start of the year and at the end, add them, divide by two. If you started with 48 guards and ended with 52, average headcount is 50. If you processed 55 separations across that year, your turnover is 110 percent.
Run it two more ways before you draw a conclusion. Split voluntary from involuntary, because firing six bad hires is a different problem from losing six good guards. Then split by site. It is common for one difficult client to be generating a disproportionate share of your departures while the rest of the book is stable.
Where in the tenure curve are you losing people?
This is the question that changes what you do next. Bucket every separation from the last twelve months by how long the guard lasted.
| Tenure at exit | What it usually means | Where to look |
|---|---|---|
| Never showed for shift one | You hired someone who was still job hunting | Time from offer to first shift, and whether anyone spoke to them in between |
| Under 30 days | The job was not what they were told it was | What the posting promised, what the first shift was actually like |
| 30 to 90 days | Onboarding or first paycheck failure | Pay accuracy, post orders, whether a supervisor ever visited |
| 90 days to a year | Schedule or supervisor | Last-minute changes, forced overtime, who they reported to |
| Over a year | Pay or a ceiling | Market rate, and whether there was anything to be promoted into |
If the mass sits in the first ninety days, that is good news. Early departures are cheaper to fix than late ones, and the fix is usually in who you hire and how you bring them on, not in a wage increase across the whole roster.
Why do guards actually quit?
Six reasons come up over and over. They are not exotic.
- The check was wrong. A guard living close to the line does not have a two-week buffer for your payroll error. One short check can end a good employee, and the reputational damage in a small local labor market outlives the guard.
- The schedule is unplannable. Posted Thursday for a Saturday shift, changed Friday. Anyone with a second job, a class, or a child in daycare cannot work like that and will leave for someone who posts two weeks out.
- Nobody talks to them. A guard on a lone post can go weeks without a supervisor visit. Silence reads as being forgotten, and it is why regular welfare check-ins do retention work as well as safety work.
- They are treated as a body, not a person. This is the one guards name in exit interviews most often and the one management believes least.
- There is nothing above them. Flat organization, no field training officer role, no site lead, no reason that year three should look different from year one.
- They did not feel safe. A bad incident with no backup, a radio that does not reach, a client site where the guard was the only one exposed. One of those ends a tenure.
Notice how many of those are operational rather than financial. That is the standard finding, and it is why a raise alone rarely moves the number. The playbook for fixing each one is a separate piece of work.
What does one departure actually cost you?
Trade estimates for replacing a guard get thrown around freely. Ignore them and build the number from your own ledger, because yours is the only one you can defend in a budget meeting.
| Line item | Where to get it |
|---|---|
| Job posting and recruiting spend | Your monthly ad spend divided by hires made |
| Background check and drug screen | Invoice from your provider |
| Licensing or registration fees you cover | State board receipts |
| Uniform and equipment issued | Cost of a full issue, less what comes back wearable |
| Training hours | Trainee wage plus trainer wage for the hours spent |
| Administrative processing | Hours from HR and payroll at their loaded rate |
| Coverage during the gap | Overtime premium paid to cover the open post until the replacement is cleared |
That last row is usually the biggest and almost always gets left out. An open post does not go unstaffed. Someone works it at time and a half, and that premium runs until the replacement is licensed and cleared, which can be weeks. Track it and you will find the real cost of a departure is well above what the recruiting line suggests. If overtime is already a sore point, the overtime problem and the turnover problem are the same problem.
What should you track every month?
Turnover is a lagging indicator. By the time it moves, the decision was made weeks earlier. Callouts, declined shift offers and a guard who stops answering the phone are the leading ones, and they are visible in your scheduling data if anyone is looking.
How do you get a useful exit interview?
Not from a form. A departing guard has no reason to fill one in honestly and every reason to write "personal reasons" and leave.
Have someone who is not their supervisor call them, a week after the last shift, when the last check has cleared and there is nothing left to lose. Three questions do the work.
- Where are you going, and what does it pay?
- When did you first start looking?
- What would have kept you here?
The second question is the valuable one. Guards usually decide to leave weeks before they resign, and the event that triggered it is still specific in their memory. Log the answers somewhere you can read them side by side. Five of them will name the same site, the same supervisor or the same pay period, and then you know what to fix.
What you pay is the backdrop to all of this, so check your rates against what the market pays in your area before concluding the problem is cultural.
Key Takeaways
- The 100 percent figure is trade shorthand. Calculate your own: separations divided by average headcount over twelve months.
- Split every separation by tenure. Departures inside ninety days are a hiring and onboarding problem, not a pay problem.
- Cost a departure from your own line items, and include the overtime premium paid to cover the open post.
- Track ninety-day turnover, median tenure, and turnover by supervisor as separate numbers.
- Exit interviews work by phone, a week later, asked by someone who was not the guard's boss.
Continue Reading
Security guard retention: the five levers
Security guard retention comes down to five levers: pay structure, schedule, supervisor, promotion rungs and the first ninety days. Pay alone will not do it.
Security guard staffing shortage: what works
The guard shortage is churn, not demand. Where the pay gap comes from, when guards actually quit, and the staffing fixes that hold a roster together.
Customer service for security companies that win
Customer service is what security clients renew on. Teach officers to greet, know the building, enforce rules by explaining them, and measure how it lands.