Roughly two out of three facility teams still run preventive maintenance out of a spreadsheet.

Sit with the number before explaining it away. In 2026 — with the smart buildings market at $141.8 billion per Grand View Research, with sensors cheap and software abundant — the operational core of most commercial maintenance programs is a file. Often one file. Often on one laptop. Frequently with a filename ending in _v3_FINAL_revised.

How the number actually happens

Nobody decides to run a portfolio on a spreadsheet. The spreadsheet wins by accretion. It starts honestly: a small operation, twenty assets, one person, and a grid of rows that genuinely works. The operation grows; the spreadsheet grows with it — new tabs, color codes, a column only Janet understands. Each individual day, the spreadsheet is good enough, and replacing it is a project nobody has time for precisely because they’re busy maintaining the spreadsheet. The cost never arrives as a single event. It arrives as a thousand small absences: the PM that wasn’t logged, the asset that was never added, the date that was overwritten instead of versioned.

And it persists because the market’s standard answer — “buy a CMMS” — fails often enough to validate the skeptics. Verified Market Research data puts roughly one in four CMMS implementations in the failure column. Most operations know one of those stories personally. The spreadsheet, whatever its sins, has never required a six-month implementation.

What most operations do when they see the 64%

Nothing — and the reasons are rational, which is exactly why the number is stable. The spreadsheet is free, familiar, and flexible. The pain it causes is chronic rather than acute, and chronic pain doesn’t trigger projects. The failed-CMMS stories provide cover. And the framing everyone uses — “we should modernize our tools” — makes the problem sound cosmetic, a matter of interface preference. So it waits.

The spreadsheet’s real cost isn’t inefficiency today. It’s that the spreadsheet destroys data at the moment of creation — and data is the prerequisite for every capability the operation will want next.

The actual failure point: the data that never existed

Here’s the reframe that changes the decision. The spreadsheet’s real cost isn’t inefficiency today. It’s that the spreadsheet destroys data at the moment of creation — and data is the prerequisite for every capability the operation will want next.

Be precise about the mechanism. A spreadsheet row says a PM was done on a date. It does not capture: who performed it, what they found, what readings the asset showed, what parts were used, how long it took, what it cost, or whether completion was verified versus merely typed. That context existed — in the technician’s hands, on that day — and the spreadsheet had no field for it, so it evaporated. Multiply by every work order for ten years: the operation has a log of dates and an institutional memory of nothing.

Now connect it forward. Predictive maintenance for commercial buildings runs on asset histories — failure patterns, condition trends, cost curves. AI scheduling runs on clean work-order data. McKinsey found only 5% of AI programs in commercial real estate achieved their objectives, and the consistent failure point is exactly this: models pointed at data that doesn’t exist or can’t be trusted. The 64% isn’t a tools statistic. It’s a measurement of how many operations are currently ineligible for the technology they’ll be sold next year.

The condition that changes it

The exit isn’t “stop using spreadsheets.” It’s making structured data capture a byproduct of doing the work, rather than a clerical task after it. Work orders that carry asset, cost, parts, findings, and verification as fields filled in the flow of execution — by the technician closing the job, by the sensor that triggered it, by the payment that released on verified completion. An asset registry built automatically from the work that touches the assets, not from a data-entry weekend that never happens. Migration that starts from the spreadsheet rather than demanding its abandonment — which is, concretely, how Sweven FM onboards operations, because demanding a clean slate is how that one-in-four failure rate happens.

STAGE 1 Data as a Byproduct

Structured data capture happens in the flow of execution, recorded instantly by the technician, sensor, or payment release.

STAGE 2 Automated Registry

The asset registry builds itself organically from the actual work that touches the assets, eliminating manual data entry.

STAGE 3 Iterative Migration

Transition starts directly from the existing spreadsheet, avoiding the clean-slate demands that cause implementations to fail.

The Core Question

The question the 64% should raise in your operation isn’t “should we get better software.” It’s this: three years from now, when you want the system that predicts failures and defends your compliance position — will the data it needs exist? Because it’s being created, or destroyed, today.

Sources: