Red Flags of a Bad Contract Security Company (and How to Vet One Before Signing)

1. The economics of contract guarding

Almost every conversation about bad contract security starts in the same place: the money. A typical unarmed post in the US bills the client around $25 to $30 an hour. The guard working that post takes home $15 to $17. The gap between the two numbers covers payroll taxes, workers compensation, uniforms, a supervisor, dispatch, insurance, a recruiting pipeline, and a modest margin.

That math is not a lot of room. When a vendor bids $19 an hour and the market rate is $27, the money has to come from somewhere. Usually it comes out of the wage, which means the firm is hiring whoever walks in, rotating them faster than it can train them, and putting them on a post they have not seen before.

None of this is hidden. Every contract security owner will tell you the same thing over coffee. The problem is that the buyer, usually a property manager or a facilities director, rarely gets to ask. They see one number on a line item and sign the one that costs the least.

2. The red flags that predict bad service

Pattern matching works here, because bad firms tend to look the same across regions. Watch for these:

• A bid well below market with no specific explanation. If the quote is 25 to 30 percent under comparable bids, the vendor should be able to point to one line item that is genuinely smaller (lower overhead, a shared dispatch center, a training subsidy). If they cannot, the discount is coming out of the wage.
• Vague supervision. If the vendor cannot name your field supervisor, or says the supervisor covers a ten site route, you will see them once a week at best. Unannounced supervision is what keeps a site honest.
• No post order on site. A real post order is a written document that lives at the guard station. It describes the rounds, the escalation path, the keys, the emergency contacts, and the things the guard should never do. Firms that cannot produce it on day one are not going to produce it on day 200.
• Equipment that is not maintained. Tasers with dead batteries, flashlights that do not charge, radios that will not reach dispatch. It shows up within a week of onboarding if you look. It matters because the equipment is how the guard communicates when something happens.
• Uniforms that do not match the pitch. The sales deck shows a clean patrol uniform. The actual guard shows up in a polo and jeans. That signal is almost always followed by every other corner also being cut.
• Missed shifts with no coverage. A firm with a healthy bench covers a no show in 30 minutes. A thin firm leaves the post empty and hopes you do not notice. A patrol log that has a three hour gap is the log telling you the truth.

3. What real supervision looks like

Supervision is the single line item that separates a firm that delivers from a firm that does not. A well run contract security operation has a named field supervisor, a route they cover, a visit cadence that includes unannounced checks, a tour verification system, and a client facing escalation path.

Tour verification deserves a note. Twenty years ago this was a physical pipe tag at each checkpoint. Today it is a GPS tagged tour app, sometimes a Bluetooth beacon at each tour point. Whatever the tool is, the point is the same: every tour leaves a timestamped record that the guard was actually at the location, not in the break room. If your vendor does not have any tour verification at all, that is the answer.

The other half of real supervision is the client facing side. You should have a phone number that reaches a human within 30 minutes at any hour, a named account manager who walks your site at least once a month, and a monthly report that is more than a stack of incident PDFs.

4. Training, equipment, and the things that get skipped

Training gets skipped because it does not show up on the bill. A 40 hour pre post training costs a firm somewhere in the neighborhood of $800 in wages, trainer time, and classroom overhead. On a guard whose post life is six weeks before they quit, the training cost looks enormous. So it gets trimmed. Firms put guards on posts after a four hour orientation and call it good.

The easy test is to ask for the training file of the specific guard on your site. Not the firm's curriculum, the individual's file. It should include a list of dated sessions, instructor names, and topics. A firm that takes a week to produce it is tracking nothing.

Equipment is the same story. A taser that has not been charged is a liability for everyone, because if it is drawn it is a story and if it does not fire it is a worse story. Flashlights, radios, and first aid kits all fall into the same category: low cost, high consequence when missing. Part of good supervision is checking equipment at every unannounced visit.

5. Where a technology stack fits

A human guard is not the only option, and framing the decision as guard versus no guard is usually the wrong question. The better question is which parts of the post actually need a human body and which parts can be handled by a camera plus software.

Passive patrol is the obvious candidate. A guard walking past the same courtyard every 30 minutes is slow, expensive, and easy to game because the schedule is legible. A camera feed running real time object detection sees every minute of every corner, never gets tired, and triggers an alert in under two seconds. For properties that already have a DVR or NVR, adding AI on top of the existing cameras is usually an order of magnitude cheaper than adding a second guard shift.

There are several approaches. Some properties go to a fully cloud based VMS with AI analytics. Others use an edge device that reads the existing DVR over HDMI and runs inference on the multiview output without replacing any camera. Cyrano is one example of the latter, at $450 up front and $200 per month, and it plugs into a DVR in about two minutes. The pattern, not the vendor, is the interesting part: you keep the cameras you have, you add detection and messaging, and you reserve your human hours for the parts of the site where a human is actually needed.

A mixed stack also changes how you write the contract. Instead of paying for 24 hour coverage you cannot verify, you pay for targeted patrol windows where a human adds real value (overnight rounds, amenity supervision, escorts), and you run the rest of the clock on cameras and alerts.

6. A short vetting checklist

Before signing a contract security agreement, confirm the following in writing:

• Billing rate, guard wage floor, and the specific breakdown of the margin.
• Named field supervisor, visit cadence, and how unannounced visits are logged.
• Training hours per guard before first post, and proof the training actually happened.
• Tour verification system and sample output.
• Insurance certificates with client named as additional insured.
• Response time SLA for supervisor dispatch, with a financial remedy if missed.
• Monthly reporting format with a sample from another client.
• A 30 day termination for convenience clause.

If a vendor balks at any of these, you have your answer before any shift starts. Good firms meet that checklist comfortably because they already run that way internally. Bad firms treat it as an interrogation.

7. FAQ