How to evaluate guard
management software
Every vendor in this category can show you a map with dots on it. What separates them is duller and far more expensive to get wrong: what the bill actually counts, what the app does when the signal drops, who has to configure the thing after the salesperson stops calling, and what you are allowed to take with you if it does not work out. This guide is the checklist we use ourselves, written so that it works on any vendor. It works on us too, and it is meant to.
Your requirements are last month, not a feature list
Before anyone shows you a dashboard, write down the four or five things that went wrong in your operation in the last thirty days. A client asked what happened at a site and nobody could say. Two officers disputed their hours and the only evidence was a signed sheet. A patrol was billed and not performed, or performed and not believed. A supervisor spent Sunday afternoon rebuilding a roster on a phone.
That list is your requirements document. Everything else on a vendor comparison is somebody else's problem being sold to you as yours. If a capability on the shortlist does not attack one of those four or five items, it is not a reason to buy and it should not be a reason to pay more.
Put every quote into your own unit before you compare anything
Vendors do not sell the same thing. Some bill per user, counting every login including office staff and the client contacts you give portal access. Some bill per guard, counting officer accounts only. Some bill per post or per site, counting the properties you cover regardless of how many people rotate through them. These are not variations on a price. They are different pricing shapes, and they reward opposite kinds of company.
Take a firm with forty officers covering twelve posts. Under a per-site model it pays for twelve, and a hiring round of twenty more officers adds nothing to the bill. Under a per-user model that same hiring round raises the software line by half. Reverse the ratio, give a firm eight officers spread across thirty small sites, and the per-site model becomes the expensive one. There is no cheaper model in general. There is only the model that suits your ratio of people to places.
So do the conversion yourself. Write down your officer headcount, your covered posts, your office staff and the number of client contacts you intend to give access to. Send that same picture to every vendor on the shortlist and ask each of them for one figure covering the first twelve months, with setup, training and every module you were shown included in it. Two of those figures are comparable. Two monthly rates are not. For a longer version of this arithmetic, with published prices from several vendors, see how much guard software costs.
Four questions that do most of the work
A full evaluation runs to a dozen items, and the table below has them. But if you have twenty minutes on a call and no more, these four separate two offers faster than any feature grid:
- What is the entire first year, in one number, in writing? Not the monthly rate. Licence, setup, training, minimum commitment and every module that appeared in the demo, added together for your real headcount.
- What does the billing unit count? Guards, users, posts or sites. Then the follow-up nobody volunteers: what a seat does when an officer leaves in week two, and whether client logins are billable.
- Show me the app with no connection. Not described. Shown, in airplane mode, on a phone, in this meeting.
- Who runs this in six months? If the answer is a specialist, you are hiring a specialist. If the answer is your own supervisor, ask them to prove it by having your supervisor add a site while you watch.
The airplane-mode test
This is the one that changes minds, and it costs ninety seconds. Ask for the demo handset, put it into airplane mode, and ask the person presenting to carry on: scan a checkpoint, open an incident, attach a photograph, close out a shift report. Then switch the radio back on and go and read what landed in the control room.
Three things are being tested at once. Whether the work was possible offline at all. Whether the record arrived carrying the time it was genuinely made, rather than the time it eventually synced. And whether anything vanished quietly, which is worse than an error message, because an error message tells the officer to try again.
The reason this matters more than it sounds is geographic. Guarding happens in basements, stairwells, plant rooms, loading bays and the far corner of a car park at four in the morning. Those are precisely the locations a client later asks about, and precisely where a platform that assumes coverage will leave you with nothing to show. A vendor who cannot do this on request has told you something, whatever they say next.
Whose clock, and can it be edited
A timestamp taken from the handset can be changed on the handset by anyone who knows how to open the settings. A timestamp assigned by the server cannot. In a billing dispute or a serious incident review, that distinction is the whole value of the record, and it is almost never on a feature grid.
Ask which clock stamps a checkpoint scan, and then ask the harder version: what the audit trail shows when a device clock is wrong, or when a record made at 02:10 syncs at 08:00. A good answer describes both times being kept. A vague answer means you will find out during the dispute.
The twelve-point checklist
| Ask this | Why it decides things | How to verify the answer |
|---|---|---|
| What is the whole first-year invoice, as one number? | Monthly rates are not comparable until setup, training, add-on modules and any minimum commitment are inside the same figure. | Ask for it in writing, itemised, for your real headcount and your real site list. A vendor who will only quote a monthly rate is quoting half a price. |
| What does the billing unit count? | Per guard, per user, per post and per site are four different bills for one company. The gap widens every time you hire. | Give both vendors the same roster and the same site list and make them price it. Compare the two totals, never the two headline rates. |
| Are office and client logins billable? | A portal you hand to twenty client contacts is either free or a line item, and nobody volunteers which. | Ask what an admin seat, a supervisor seat and a client contact each cost, and whether there is a floor below which you pay anyway. |
| What happens to a seat when an officer leaves? | Turnover in guarding is high enough that mid-month joiners and leavers move the bill on their own. | Ask whether the count is a month-end snapshot, an average or a peak, and whether deactivating an account this week credits anything. |
| Does the app work with no connection? | Posts include basements, stairwells, plant rooms and loading docks. Software that needs bars produces gaps in exactly the places you get disputed. | Airplane mode, during the demo, on a real handset. Scan, report, photograph, then reconnect and read the timestamp that arrives. |
| Whose clock stamps the record? | A time taken from the handset can be changed on the handset. A time taken from the server cannot. | Ask directly, then ask what the audit trail shows if a device clock is wrong or a record syncs six hours late. |
| Is checkpoint hardware required? | Wands and readers carry purchase, breakage, replacement and the labour of collecting and downloading them, for as long as you run the contract. | Price five years of it, including lost units, and compare against tours that run on the officer’s own phone with a printed code. |
| Who administers this once the salesperson has gone? | A platform that needs a configuration specialist needs one permanently, and that role is rarely in the quote. | Have your own supervisor, not the vendor, add a site, build a rotation and change a patrol route while you watch the clock. |
| Is the client portal in the base price? | Client-facing reporting is often the reason for the purchase and often the module that is priced separately. | Get the answer in writing, and ask whether the client receives a live portal, an app, or a PDF someone emails on Monday. |
| What language reaches the officer? | A localised marketing site says nothing about the language of the menus, the incident types or the monthly report. | Ask to see a full month’s client report and the guard app in the language your people work in, and check the vocabulary against the words your supervisors actually say. |
| What can we export if we leave? | Your evidence is your defence in a dispute. If it cannot leave, you are renting your own history. | Ask for a sample export before you sign, in a named format, and ask what it costs to obtain it after the contract ends. |
| What does the contract do at renewal? | Auto-renewal windows, notice periods and uplift clauses decide how expensive a mistake is. | Read the renewal and termination clauses before the feature list. Ask for the notice period in days and the maximum annual increase. |
This table names no vendor and makes no claim about any product. It is a list of questions, and the middle column explains why each one is worth the time. Put all twelve to every company on your shortlist, CGuardPro included, and keep the answers in writing.
What changing systems actually costs you
The technical build is smaller than most buyers fear. Creating clients, sites and posts, importing a roster, drawing patrol routes and printing checkpoint codes is a few days of work for a firm of ordinary size, and much of it is data entry that a vendor will do for a fixed fee if you would rather not.
The cost that surprises people is habitual. Somebody has to decide which incident categories you will actually use, which post orders need rewriting because they referenced a paper book, who holds which permission, and how you will stop supervisors keeping a private second record in a notebook or a group chat because that is how they have always done it. That work is not billable to anyone and it runs for weeks.
Three practical rules. Start with one client and two or three sites rather than the whole book, because a small deployment that visibly works is a far better internal argument than any presentation. Establish what your current provider will release before you give notice, since the outgoing vendor's export options set the ceiling on what comes with you and not the incoming vendor's import options. And do not schedule a migration during a major client's peak season, an audit window, or the fortnight either side of a contract renewal.
How to spend the trial
A trial spent exploring an empty demo account is a waste of two weeks. Load one real client, two or three real sites, and the genuine rotation with its genuine gaps. Print the codes. Put the app on the phones of officers who did not ask for it and will not be flattered by being consulted.
Then measure three things at the end. How many records were filed without anyone being chased. Whether the difficult posts produced as much evidence as the easy ones, or whether the basement site went quiet. And how many hours of your supervisor's week the system consumed rather than saved. Those three numbers predict the next three years more reliably than anything in a proposal.
Frequently asked questions
How do you evaluate guard management software?+
Work in this order and it takes about two weeks. Write down the four or five things that went wrong in your operation last month, because those are your requirements and nothing else is. Shortlist three vendors, no more. Send all three the same roster and site list and ask for a single first-year figure in writing. Run each demo on your own sites rather than a sample account, and insist the guard app is shown on a phone in airplane mode. Then take the two finalists into a trial on one real client with two or three sites, and judge them by whether officers keep using them in week two without being chased. A feature grid will not tell you that; a fortnight of real shifts will.
What should be on a guard management software checklist?+
Twelve items cover almost everything that later turns into a complaint: the total first-year cost as one figure, what the billing unit counts, whether office and client logins are billable, what happens to a seat when an officer leaves, offline behaviour, whose clock stamps the record, whether proprietary checkpoint hardware is required, who administers the system after go-live, whether the client portal is included, what language the officer and the client actually see, what you can export if you leave, and what the contract does at renewal. The table on this page gives each one alongside why it matters and how to verify it rather than take an answer on trust.
How do you compare prices when vendors bill differently?+
Convert everything to your own annual total before you compare anything. Take your real officer headcount, your real number of covered posts, your office staff and the client contacts you intend to give access, then ask each vendor to price that exact picture for twelve months including setup and training. A firm with a lot of officers rotating through a few posts and a firm with thin staffing spread over many small sites will find different billing units favour them, so there is no universally cheaper model. There is only the cheaper model for your ratio, and you cannot see it from a pricing page.
What is the airplane-mode test?+
You put the demo handset into airplane mode and ask the salesperson to keep going: scan a checkpoint, write an incident, attach a photo, close the shift report. Then you turn the radio back on and read what arrives in the control room. What you are checking is whether the record kept the time it was actually made, whether the photo survived, and whether anything was silently dropped. It takes ninety seconds and it is the single most informative thing you can do in a demo, because guarding happens in the places where signal does not.
How long does it take to switch guard management systems?+
The build is short and the habit is long. Creating clients, sites and posts, loading a roster, drawing patrol routes and printing checkpoint codes is a matter of days for a firm of normal size. What runs into weeks is the human part: agreeing incident categories, rewriting post orders, settling who holds which permission, and getting supervisors to stop keeping a private second record in a notebook or a group chat. Plan for the routine to feel automatic at every site somewhere around the second month, and do not start during a major client’s peak season or an audit window.
Can we take our historical data with us?+
Master data moves cleanly. Clients, sites, posts and staff are structured records and any competent vendor will import them. Historical evidence is the harder question, and the limit is set by your outgoing vendor rather than your incoming one: what patrol logs, incident records and photographs you can extract, in what format, at what price. Ask your current provider for a sample export while you are still a customer in good standing. Asking after you have given notice is a much weaker position, and the answer sometimes changes.
How long should a trial be, and how should we spend it?+
Two weeks of real work beats a month of clicking. Pick one client and two or three sites, load the genuine roster with its genuine rotation, print the codes, and let officers run ordinary shifts on their own phones. Then look at three things: how many records were filed without anyone being reminded, how many came from the difficult posts rather than the easy ones, and how much of your supervisor’s week the system consumed. Adoption at the post decides whether the platform produces data at all, and no demo predicts it.
What does CGuardPro cost, and does this checklist apply to you?+
It applies to us and we would rather you used it. CGuardPro publishes $5 per user per month, where admins, guards, supervisors and client logins all count as users, plus an optional one-time $299.99 implementation package and a 14-day free trial. That is enough for you to build a first-year figure before you speak to anyone. Every other question on this page is one we expect to answer in a demo, in airplane mode, on a real phone, with your sites loaded rather than ours.
Put the twelve questions to us
Bring your roster, your site list and the four things that went wrong last month. We will answer all twelve on the call, and the offline one on a handset.