Non-billable overtime: why guard contracts lose margin.

A guard calls off at nine at night. The post cannot go dark, so the officer already on site stays into a tenth and eleventh hour. Payroll goes up tonight. The invoice to the client does not. That gap, between the hours you pay for and the hours you can bill, is where contract security margin quietly drains away.

It isn't an occasional accident. It's built into the deal. Most guard bill rates are locked in the client contract. They can't be adjusted when a shift runs into overtime. So the premium you pay to keep a post covered lands on your own books, not the customer's. Industry trade press has long estimated that over 85% of client agreements don't allow you to bill overtime, and operators are still working on it: a state guarding association ran a 2024 feature on navigating overtime law and cutting non-billable hours.

So the question worth answering isn't whether non-billable overtime costs you. It's where it starts, and what actually closes the gap. Here's how the money leaks, and where the fix lives.

In this post:

• Why isn't security guard overtime billable to the client?
• How much does non-billable overtime actually cost?
• What actually causes runaway guard overtime?
• How do you reduce guard overtime without leaving posts unmanned?
• What does it take to run coverage that protects margin?

Why isn't security guard overtime billable to the client?

In most contract guarding, overtime isn't billable because the bill rate is locked in the day the contract is signed, and it can't be reopened when your costs climb. Most businesses can pass a cost increase down the line to the customer. Contract guarding usually can't.

A guard bill rate is built from three things: the pay rate, overhead, and a profit margin on top. Firms typically mark the wage up 1.5 to 2 times to get there. In one 2025 example, a guard billed at $32 an hour against a $19.13 wage left about a 12% net margin on that post.

Here's the part that matters. The bill rate is set for a standard hour. So when you cover that same post on overtime, you pay time and a half on hours you already sold at the standard price. The extra half comes out of that thin margin. You aren't just earning less on those hours. You're paying out of pocket to keep the post covered.

How much does non-billable overtime actually cost a security company?

The industry has a name for it: non-billable overtime, or NBOT. That's overtime a firm pays but can't invoice to the client. Estimates put it at 7% to 11% above what a firm pays for its scheduled hours, which works out to a 6% to 10% hit to profit.

Now hold that number up against the margins. Contract guarding is a thin-margin business. Gross margins usually run in the mid-teens to low twenties. After overhead, the operating margin sits around 10%, and lower for firms competing on price. A profit hit that size against a margin that thin isn't a rounding error. For some contracts, it's the line between a post you make money on and one you're paying to keep.

This isn't time theft, which is a different problem. Buddy punching and padded timesheets are real, but that's fraud, and it has its own fixes. Non-billable overtime isn't fraud. It's work that really happened, that really had to happen, and that the contract simply won't pay you back for.

What actually causes runaway guard overtime?

The real cause of guard overtime is the coverage gap, not a loose approval policy. An open post has to be filled right now, so the officer already on duty stays into overtime. And constant turnover keeps reopening those posts, which is why tightening approvals alone rarely moves the number.

When finance sees overtime climbing, the instinct is to clamp down on approval. Require a manager sign-off. Flag anyone nearing forty hours. Treat the overtime line as the problem. It feels like the obvious place to push, because that's where the cost shows up.

It rarely works, because overtime is the symptom, not the cause. When a post opens with no one cleared to fill it, saying no to overtime isn't really an option. The alternative is a dark post, a broken service level, and a client asking why the lobby was empty at midnight. So the overtime gets approved, every time.

And the gap keeps reopening, because the workforce keeps turning over. Annual guard turnover runs between 100% and 300%. As of 2025, more than 40% of security providers call turnover their single biggest challenge. Every guard who leaves is another post to re-cover, usually on short notice, often with someone already on the clock.

The timing makes it worse. Federal law sets overtime at time and a half after forty hours in a week, and some states add daily overtime on top. A few coverage scrambles in one week can push a guard past the threshold fast. From there, every extra hour carries a premium the contract won't pay.

How do you reduce guard overtime without leaving posts unmanned?

You reduce guard overtime by closing the coverage gap before it opens, not by policing overtime after it's logged. The move is to fill each open post with a guard who can actually take it, before overtime is the only option left.

So if approval isn't the lever, what is? It's the question of who fills the open post, and whether you can answer it fast enough.

The right guard for a sudden gap is a specific person. They're trained and certified for that site. They're close enough to get there in time. They aren't about to cross a weekly or daily overtime threshold. They aren't about to hit a rest limit, the rules that cap how many hours someone can work in a row. And they're allowed to take the shift under that site's union contract. Find that person fast, and the gap closes at the standard rate. Miss them, and it falls to whoever is already standing there, on overtime.

At a single site, a scheduler can keep all of that in their head. But across dozens of sites, several states, and a handful of different union contracts, the question "who can take this post without tripping a premium or breaking a rule" has too many moving parts to answer from memory at nine at night. So the default takes over. And the default is overtime. The decision that costs you the margin happens when the gap opens, not when the overtime gets approved.

That is the real margin lever, and it is not a matter of trying harder. It is a matter of what you can see the moment a gap opens.

What does it take to run coverage that protects margin?

What it takes is a system that makes eligibility visible in real time: one that knows, the instant a post opens, which guards can take it cleanly and which would cost you a premium or break a rule.

Getting that right across a large, unionized, multi-site operation is a genuinely hard problem, and it's the one WorkAxle was built to solve. When a post opens, it puts the eligible guards in front of the scheduler and keeps everyone else off the list, so an open-shift offer can't go to someone it would push into overtime or a rule violation. Approval screens show each guard's hours and overtime impact before anyone is assigned.

It doesn't quietly block a manager who needs to make a judgment call. It warns, shows the rule in play, and lets the manager override with the full picture in front of them. The goal is to make the margin-protecting choice the easy one, not to take the decision away.

This is exactly the environment where it proves out. A national guarding firm went live 4 to 5 times faster than legacy, running several union contracts across major Canadian airports from one system. That's where coverage gaps pile up, and where most scheduling tools quietly give up and let overtime cover the difference. Closing those gaps in the moment, instead of paying for them later, is what protects the margin the contract promised.

What this means for your margin

Stop measuring overtime only at the approval line. By the time a manager signs off on an overtime hour, the margin decision was already made, hours earlier, when a gap opened and no one could fill it cleanly. The approval is the receipt, not the decision.

Look further back. Where do your coverage gaps start? How often does a call-off or an open post have no qualified, available guard to take it, so it defaults to overtime? That number, not your overtime approval rate, is where your margin lives or dies. It's also the number most reports never break out.

If you want to see where your own coverage model is quietly forcing overtime, that's something you can measure. And it's the first step to billing for the hours you actually work.