Guard turnover is the industry’s chronic cost. Plenty of contract firms replace something close to their whole roster inside a year, and every departure carries the same bill twice over: what you spend to recruit, license and train the replacement, and what you lose while a post runs on overtime or an unfamiliar officer in the meantime.
Before you try to fix it, calculate your own number — separations over the period divided by average headcount — and price a single replacement using your real recruiting, licensing and orientation hours. The industry averages you’ll find quoted online are worth less to you than your own two figures.
Here’s what actually moves them.
Why Guards Leave
Run exit interviews for two quarters and the same reasons surface in nearly every guard force. Roughly in the order they tend to appear:
- Pay, especially where a competitor two miles away pays a dollar more
- Unpredictable schedules — the roster posted late, or changed after it’s posted
- Feeling invisible — no one from the office has been to the post in months
- No path upward, so the good officers leave first
- Payroll disputes, particularly over unrecorded overtime
- Post conditions — no relief for breaks, no working lighting, no radio coverage
Your order will differ, and that’s the point: fix the ones your own exit interviews name, not the ones an article ranked for you.
Strategy 1: Fair and Transparent Scheduling
An unpredictable roster is the reason officers give most often after pay, and it is the cheapest one to fix. An officer with children, a second job or a commute cannot arrange any of it around a schedule that lands on Friday for a Monday start.
What works
- Publish the roster on a fixed cadence and hold to it. The interval matters less than the fact that it never moves. Two weeks out is a common target; a reliable one week beats an erratic three.
- Collect availability and preferences once, then honor them. Officers stop submitting preferences the second they learn nobody reads them.
- Let officers trade shifts inside rules you set — same qualification, no overtime created, both parties confirm, and the swap written down before you allow it.
- Distribute overtime openly. Overtime that always lands on the same three people reads as favoritism to everyone else, whether or not it is.
- Avoid split shifts wherever the contract allows, and pay a differential where it does not.
Where software helps, and where it doesn’t
A scheduling system does not make a roster fair. What it does is make the roster visible: officers see their own upcoming shifts on their phone instead of waiting for a text, changes reach them immediately, and the running hour totals that drive overtime distribution are in front of the scheduler while they are still assigning rather than after payroll closes. The fairness policy is still yours to write. The tool only makes it enforceable and auditable.
Strategy 2: Close the Distance to the Post
Officers work alone, at hours when the office is closed, often for a client whose staff barely acknowledges them. “Nobody from the company has been here in months” is one of the most common lines in an exit interview, and it costs nothing to fix.
- Set a minimum site-visit frequency per supervisor and track whether it happens. If you cannot say when a supervisor last stood on a given post, that post is at risk.
- Make one contact per officer per week that is not a schedule change or a correction. A brief message that acknowledges a specific thing they did is enough.
- Answer field messages within a stated window, and tell officers what that window is. Unanswered questions from a post at 3 a.m. teach people the company is not there.
- Run recognition as a standing process rather than an impulse. A structured recognition program survives a busy quarter; goodwill does not.
- Give officers a channel to raise post conditions — broken lighting, no break relief, a client contact who is abusive — and close the loop on what you did about it. Conditions you never hear about become resignations you never predicted.
Strategy 3: Provide a Path Upward
“Security officer” does not have to be a terminal job, and the officers who believe it is are the first ones a competitor takes.
A workable ladder
- Security Officer — entry level
- Senior Officer — proven on post, trusted with the difficult sites
- Shift Lead — responsible for a relief and a passdown
- Site Supervisor — owns one property and the client relationship on it
- Area Supervisor — several sites, the roster and the escalations
- Operations Manager — company level
The ladder only retains anyone if the criteria for each rung are written down and the promotions visibly follow them. A ladder that exists on paper while promotions go to whoever the manager knows is worse than no ladder at all.
Training that officers value
- Cross-train on different post types so people are not stuck on one site forever.
- Sponsor the commissioned (armed) credential where your state allows it and your contracts need it — this is one of the few investments an officer can carry with them, which is exactly why it buys loyalty.
- Offer first aid and CPR certification.
- Give supervisor candidates actual management training before you promote them, not after they have already lost two people.
Whatever you offer, put it in a structured training program rather than handling it case by case, so it survives a change of staff in the office.
Strategy 4: Get Compensation Right — With Your Own Numbers
You do not have to be the highest payer in your market. You do have to know where you sit, and most operators do not.
Build your own benchmark instead of trusting a published rate table: collect the posted rates in the job ads competing for your officers in each city you operate in, separate armed from unarmed and standing from patrol, and refresh it every quarter. Rates move by market and by post type, and a number printed in an article is stale before you read it. Our pay benchmarking guide walks through the collection method.
Then check the other half of the equation. If your pay rate is below market, the question is whether your bill rate can carry an increase, or whether you signed the contract too thin to begin with. Raising pay you cannot bill for is a slower way to lose the account.
Beyond the hourly rate
- Attendance and reliability bonuses, sized against what a single uncovered shift actually costs you
- Referral bonuses paid on a tenure milestone rather than on the hire date
- Any benefits contribution at all, which still differentiates in this market
- Uniform allowance, and a phone stipend if officers use personal devices for work apps
Also audit your payroll for accuracy before you spend anything on incentives. Disputes over unrecorded overtime and missing hours show up in exit interviews far more often than the base rate does, and they are the cheapest complaint on this list to eliminate.
Strategy 5: Fix the First Ninety Days
Track how many of your hires are still on post at ninety days. In most guard forces that number is the earliest and loudest warning you get: when people leave inside the first quarter, the cause is almost never pay — it is selection, orientation, or a first week nobody planned.
First week: an experienced officer alongside them rather than a solo start, written post orders instead of a verbal briefing, equipment that works and fits, a face-to-face introduction to the supervisor who will actually manage them, and their app set up and tested before the shift starts rather than during it.
First thirty days: a weekly check-in, a deliberate second site so they are not locked to one post, a formal conversation at day 30, and immediate handling of anything raised in it.
What to Track Monthly
- Turnover rate broken out at 30 days, 90 days and twelve months — one blended annual figure hides the problem
- Reason for departure, categorized, not filed as “personal”
- Retention by site: some posts are consistently where people quit
- Retention by supervisor: management quality varies more than most owners admit
- Cost per hire, so retention spending has something to be compared against
The Bottom Line
Do this arithmetic with your own two numbers rather than anyone’s benchmark. Take your separations over the last twelve months divided by average headcount, and price one replacement using your real recruiting, licensing, orientation and overtime-coverage hours. Multiply the hires you would avoid by that cost. Whatever comes out is the budget you are actually arguing about — and unlike an industry average, it is a figure you can defend to an owner who asks where the number came from.
Retention work competes for that budget against everything else. The reason it usually wins is that the same changes — predictable rosters, clean payroll, a real first week — also reduce the coverage gaps and payroll disputes that cost you separately.