Looking for a CMMS: where to start

When someone starts looking for a maintenance program, they usually do it backwards: looking at products before deciding what they need. Since every spec sheet says the same thing — work orders, preventive maintenance, assets, reports, mobile app — the comparison doesn’t discriminate, and the decision ends up being made on price or on whoever pushed hardest.

Here’s a more useful way to approach it.

Step 1: decide the scope before looking at anything

This is the decision that shapes every other one, and the one most often skipped.

Do you want to just log technical work and pull reports? Or do you want the system to cross departments, so certain steps reach purchasing, administration, or accounting on their own without going through the manager?

A narrow scope gets implemented in weeks. One that crosses several departments requires prior agreements on who does what, and those agreements take longer than configuring the software.

Step 2: write down your three real problems

Not a wish list of features: the three concrete problems that brought you here. They usually belong to this family:

• Inspections get missed and nobody notices until there’s a breakdown.
• Stock never adds up and duplicate purchases show up.
• Hours get logged from memory, so you don’t know what anything actually costs.
• A client asks about a job report from months ago and it has to be dug up.
• The technician who left took what they knew about a site with them.

With three problems written down, every demo turns into a concrete test instead of a presentation. There’s a fuller inventory in the gaps a CMMS should fix.

Step 3: the criteria that actually set things apart

These four show up six months in and are almost never on the spec sheet.

How you pay. The per-user model looks cheap and gets complicated the moment the team grows or you need to give access to clients and subcontractors. When every seat costs money, the temptation to share accounts shows up, and as soon as two people sign in with the same user the history stops being worth anything as evidence. In GMAO CLOUD, licenses are unlimited across all three plans. Do the math with your headcount three years out.

Whether the technician will actually use it. The system is fed by whoever is out in the field. If the app is awkward or requires signal, logging gets pushed to the end of the day and stops being accurate.

What happens to old versions. A major version change can leave whoever stays behind without support, with a migration nobody budgeted for. Ask what happened last time and who bore the cost.

How your data comes out. The history is yours. If there’s no full export or API, the system ends up being expensive to leave.

There’s an extended guide on how to choose the best CMMS.

Step 4: look at the forms, not the feature list

Concrete advice that saves headaches. In the demo, ask to see the actual screens: the work order, the incident, the purchase order, the lists, and the filters.

That’s where you see whether the system fits how you work. A program that forces six fields you don’t need on every job report multiplies that friction across every order of the year.

Step 5: check the configuration cascade

A question that sets things apart and almost nobody asks: if you have two hundred identical pieces of equipment, do you have to configure them one by one?

In GMAO CLOUD the checklist template gets linked to the asset, to its model, or to an entire family, and checklists resolve in cascade, so you define it once and only fine-tune where needed. The difference between that and configuring equipment one by one is weeks of work.

Step 6: ask about integration in detail

“Do you integrate with my ERP?” doesn’t cut it: the answer is always yes. What matters is what data travels and in which direction.

Not every connector does the same thing: some bring in master data — clients, addresses, items — and don’t return orders, and vice versa. The integrations catalog details the scope of each one, and it’s worth demanding that same level of detail from any vendor. If your system isn’t on any list, what matters is that generic paths exist: REST API, SOAP, database access, CSV, and FTP.

Step 7: test it on your data, not theirs

This is the step that best predicts the outcome. With two or three of your own assets, their checklist templates and frequencies, in half an hour you see what two hours of presentation won’t show you.

And have three profiles present: the maintenance manager, who looks at planning; a technician, who spots in two minutes whether the app is going to waste their time; and whoever answers the phone, who really knows how work comes in today.

The mistakes that get expensive

Starting with the full inventory. Six months loading data before seeing any result wears any team out. Start with the critical assets.

Configuring it without the people who’ll use it. A system designed in a meeting room collides with reality on day one in the field.

Looking for a “legal maintenance module.” In GMAO CLOUD it doesn’t exist as such: it’s handled through the preventive maintenance mechanism, by linking the checklist template the regulation requires and its frequency. Be wary of anyone who promises the software guarantees compliance with a standard: it records and demonstrates, but the company is what complies.

Expecting it to fix the process. If today nobody decides who handles what or with what priority, the system will just record the same disorder with more precision.

What to do this week

Two things that don’t depend on any vendor and that hold up whichever one you choose afterward: write the list of your critical assets — the ones that stop production or service if they fail — and, for each one, what gets done to it and how often. It’s the work that takes the longest in an implementation, and the only one you can get ahead of without buying anything.

Once you have that, request a demo on those specific pieces of equipment. It’s half an hour and will tell you more than any comparison: you’ll see the preventive plan being generated with your frequencies, the order closing out with your checks, and the history starting to be written with your data — which is exactly what you’ll be using every day.