Getting maintenance reports without the work

“Reports in one click” is a promise made on every product spec sheet, and it’s only half true. The click is the easy part. What determines whether the report is any good is what happened three months earlier, when someone recorded — or didn’t record — what they did.

This article is about both things: what it takes for reports to generate themselves, and which ones are worth looking at.

Where the data comes from

A report doesn’t create information, it organizes it. If hours are jotted down from memory at the end of the day, the hours report will be a reconstruction; if material is logged on Friday, the cost per piece of equipment will be incomplete.

What makes reports trustworthy is recording happening where the work happens: stopwatch inside the work order itself, material consumed against the warehouse, and checklist filled in, from the app, which works without coverage.

Once that’s done, the report really is one click. Without it, no software can fix it.

What’s available

GMAO CLOUD has more than sixty reports with filters and export, plus a dashboard. Grouped by what they answer:

Cost. Cost per piece of equipment, equipment costs, materials used in job sheets, hours per client and time entries.

Reliability. MTBF and MTTR, downtime, intervention times per asset, anomalies.

Plan. Annual preventive plan, order log, orders by month.

Personnel. Hours per technician, non-effective hours, shifts, absences, shift summary and per team leader.

Assets. Inventory, movements, counters, traceability and action log.

Quality. Quality reports, diagnosis and repair, incidents per site and job sheets per client.

The four you need to look at

With sixty reports, it’s easy to look at none of them. If you have to pick, these four and no more:

Preventive-to-corrective hours ratio. The one that best describes whether the operation is proactive or reactive. It takes months to move, and that’s exactly why it’s the most honest one: it can’t be dressed up in a single quarter.

Accumulated cost per asset. The one that decides whether a machine gets repaired again or replaced.

Deviation between estimated and actual time. The one that surprises people most the first time, because it almost always reveals a type of job being quoted below what it costs.

Annual plan executed. The honest check on whether the plan is being met or just drawn up. And the one shown when there’s a regulatory obligation, along with the checklists and the dated documentation.

How often to check them

You almost never need a daily report. What works is one hour every quarter, with those four in front of you, out of which come specific decisions: which schedules to move up or down, which equipment concentrates anomalies, which assets have accumulated a cost that no longer justifies repairing them, and which contracts need renegotiating.

It’s the first thing that disappears from the calendar once the day fills up with urgent matters. Adding it to the calendar with its own frequency, like any other task, is a silly trick that works.

Asking in plain language

There’s an additional route worth mentioning without overselling it. Among GMAO CLOUD’s artificial intelligence features is the ability to ask about your figures in plain language: orders, hours and costs aggregated for whatever period you ask about, answered in a sentence instead of a table.

Two things matter more than the feature itself: it proposes, it doesn’t execute — it answers and suggests, it doesn’t modify anything — and every AI feature has its own on/off switch, with most of them off by default. It’s covered in detail in AI in GMAO CLOUD.

And the usual caveat: a question about an empty history has no answer.

If you want your own dashboard

Inside the product there’s a catalog of reports with filters and export, and a dashboard. What there isn’t is a visual builder for putting together new views by dragging fields around.

Instead there’s something that in practice solves that case better: connecting your own BI tool. If you already work with Power BI, Looker Studio, Tableau or Qlik, there’s a public REST API and direct SQL connectivity, both in the integrations catalog.

The advantage of approaching it this way is that your dashboard can cross the CMMS data with production data, ERP data or billing data — exactly what a builder locked inside the product could never do.

How to read them without drawing the wrong conclusions

Three rules that avoid most interpretation mistakes.

Look at the family, not the total. If you’ve changed the frequency for a type of equipment, compare that type. The site-wide total has too much noise to attribute anything to a specific change.

Give it one or two cycles. A frequency change takes time to show up. Measuring after three weeks and concluding it didn’t work is the most common mistake.

Note the date of the change. It sounds obvious and almost never gets done. Without it, before and after can’t be separated, and any reading of the report is just a guess.

And a fourth, about the data itself: check what valuation method each warehouse uses — average cost, actual cost, FIFO or LIFO — because it determines exactly what that cost-per-asset figure you’re looking at actually means.

The mistake that makes a dashboard useless

Putting thirty figures on it. Nobody looks at it.

The usefulness of a report isn’t in how much data it shows, but in the specific decision it enables: repair or replace, raise or lower a frequency, renegotiate or keep a contract. If a figure doesn’t change any decision, it’s excess.

And a warning about how to use it

Personnel reports are for sizing teams, budgeting and distributing workload. Not for watching anyone. A team that senses the system exists to keep tabs on them stops feeding it accurately, and at that point every report — including the cost ones — stops being worth anything.

Where to start if you’ve just rolled it out

The reasonable objection: all of the above assumes history, and if you’ve only had it running a month, you don’t have any.

Start with the one thing that doesn’t depend on history: the percentage of plan executed. From the first month you know how many of the generated preventive orders were closed on time, and that figure alone already drives decisions. If it’s well below target, the problem isn’t the frequency: it’s that the plan is bigger than the team can execute.

After three months, the first aggregated anomalies show up. After a year, MTBF starts to mean something. And after two years, you can compare full periods against each other.

In the meantime, what matters is that the recording is accurate from the start: the data from the first few months is what later serves as the baseline, and it can’t be reconstructed.

If you want to see what would come out with data similar to yours, you can request a demo.