Most guard contracts are bought the wrong way round. A buyer decides how many guard-hours they can afford, then asks what those hours can cover. The result is a post that is present but not positioned, or a patrol frequency that looks like coverage on paper and is predictable enough to plan around in practice.

The better sequence is to name the loss you are trying to prevent, then choose the posture that actually prevents it. There are three basic postures, and they do genuinely different jobs.

Standing post

A guard fixed at one location — a lobby, a gate, a dock, a single controlled door.

What it is good at: access control and deterrence at a specific point. If everyone entering has to pass one place, a standing post converts that place into a decision point. It is also the only posture that produces a consistent record of who came and went, because the same person sees every entry.

What it does not do: cover anything out of sight of the post. A standing post at the front of a twelve-acre site does nothing at all for the back fence, and buyers regularly discover this after the loss.

Standing post is the right answer when the risk is concentrated at a boundary — a single entrance, a cash-handling point, a reception area where staff and the public meet.

Mobile patrol

A guard moving through an area on a route, either on foot within a site or by vehicle across several.

What it is good at: covering ground that cannot justify a permanent body. A patrol touches many points cheaply. It creates uncertainty for anyone watching the site, provided the route and timing vary.

What it does not do: guarantee presence at any given point at any given time. A patrol is a sampling method. If your intruder needs eleven minutes and your patrol returns every forty, the patrol is not a control — it is a witness that arrives afterwards.

The failure mode is predictability. A patrol that runs the same route at the same times is a published timetable. Randomised route order, randomised interval, and checkpoint scanning that proves the route was actually walked are what separate a patrol from a drive-past.

Patrol is the right answer for large, low-density ground: construction sites, industrial yards, storage, multi-building campuses, vacant property. For very large or remote land it pairs with technology rather than replacing it — see estate and ranch security for how that works over acreage.

Armed coverage

Armed is not a third level of the same service. It is a different risk decision with a different legal frame, and it should be chosen on the nature of the threat rather than on the value of what is being protected.

Arizona sets the training floor in A.R.S. § 32-2632: at least sixteen hours of initial firearms instruction, completed before assignment to any post requiring a firearm, plus eight hours of annual continuing firearms instruction. That sits on top of the eight hours of pre-assignment training every guard receives and the eight-hour refresher tied to renewal. The duty to provide and record it falls on the licensed agency.

The decisive point is the use-of-force standard, and it is not what most buyers assume. Arizona does not justify deadly force to protect property. Sections 13-407 and 13-408 permit ordinary physical force to stop a trespass or a theft, but both route deadly force back through the self-defence provisions of section 13-405, which require a reasonable belief that deadly force is immediately necessary against another’s use or attempted use of unlawful deadly force.

So arming a post to protect metal, inventory or equipment adds liability without adding a lawful option. Arming a post where people and risk genuinely intersect is a different matter, because the justification the statute recognises is actually present. That distinction is worked through in more detail in our piece on armed security in Arizona.

The questions that actually decide it

Before comparing quotes, answer these:

What is the loss? Theft of material, unauthorised access, violence against staff, liability exposure, or a contractual requirement someone else imposed. These call for different postures.
When does it happen? If every incident is between midnight and 4 a.m., coverage from 6 p.m. to 6 a.m. is paying for eight hours that do nothing.
How long does the adversary need? This sets the maximum acceptable patrol interval. It is the single most useful number in the whole exercise and almost nobody calculates it.
Are people present? Sites with staff or public present have a violence exposure that empty sites do not, and that is what should drive the armed decision.
What does the guard do when something happens? Observe and report, intervene, or call. If this is not written down before the contract starts, it will be decided by one person under pressure at 3 a.m.

Mixed postures are usually the right answer

Real sites rarely fit one category. A distribution centre might want a standing post at the driver gate during operating hours, a patrol overnight covering the yard and the trailer line, and no armed element at all — with cameras and lighting doing the work between patrol passes. A medical building might want the opposite: an unarmed presence at reception during the day where the confrontation risk is, and no overnight presence at all because the building is genuinely empty and alarmed.

Designing that mix is a short exercise, but it has to happen before the hours are priced rather than after. Verify any agency’s credentials directly with the Arizona Department of Public Safety licensing unit — both the agency licence and the individual guard registrations.

Honeybadger Solutions scopes coverage from the risk rather than from an hours budget, and will say when a site does not need the posture it has asked for. For sites where the exposure is loss in transit rather than on the property, the discipline is different again — that is covered under transportation and cargo security.