What Maintenance Software Does

Operators reach for software first and discover afterwards that the difficult parts are not the parts software does.

What it genuinely provides

A record that persists. Asset history that survives staff turnover, vendor changes and lost spreadsheets. This is the main thing and it is worth a great deal.

Routing. Reports reach the right place without depending on someone's inbox.

Scheduling. Preventive tasks generated automatically, with reminders and completion tracking.

Aggregation. Cost per asset, per site, per vendor, across an estate, without manual compilation.

Visibility. Open work orders, ageing, overdue preventive tasks, in one view.

Document storage attached to assets — manuals, warranties, certificates, photographs.

Structure. Forms with required fields produce better data than free text, which is a real improvement in report quality.

Audit trail. Who did what, when.

What it does not do

Create the asset register. Someone walks the sites with a clipboard. This is the largest single task and no product performs it.

Improve report quality by itself. A structured form helps; a manager who does not know what to observe still reports poorly. Training does that.

Make contractors complete close-outs. Contract terms and enforcement do that.

Negotiate agreements or set priorities.

Decide repair or replace. It supplies the data for a decision someone makes.

Fix a broken process. Software applied to an undefined process produces an expensive undefined process.

Prevent failures. It supports a schedule; the technician's work prevents failures.

When it is worth buying

Site count is the main driver. A single restaurant does not need it; a spreadsheet, a vendor list and calendar reminders cover the requirement.

The transition is roughly where the operator can no longer hold the estate in their head — commonly somewhere between five and fifteen sites, depending on how involved the owner is.

Other triggers: a compliance requirement for documented records, franchisor or insurer demands, multiple people needing the same view, or an estate old enough that replacement planning matters.

The wrong trigger is a demonstration. These products demonstrate well because the demonstration data is complete, which is exactly what you do not have.

The order that works

Asset register first, in a spreadsheet.

Process second — priorities, reporting form, approval thresholds, vendor agreements.

Run it manually for a quarter. This finds the problems in the process while they are cheap to fix.

Then buy software and import what you have.

Operators who buy first spend the first year using an expensive system as a spreadsheet, and frequently conclude the software failed when the missing piece was the register and the process.

The spreadsheet version

For an operator not ready to buy, a spreadsheet does more than expected.

Sheet one: assets. Tag, site, type, make, model, serial, install date, warranty expiry, vendor, ownership.

Sheet two: work orders. Date, site, asset tag, priority, fault, action, vendor, labour, parts, total, downtime, consequential loss.

Sheet three: schedule. Asset, task, frequency, last done, next due.

Two pivot tables for cost per asset and per site.

This produces every decision described elsewhere on this site. Its weaknesses are concurrency, mobile access and reminders — real limitations, and not reasons to delay starting.

What software cannot rescue

No asset register — the system is empty.

No close-out discipline — the reports are precise and meaningless.

No vendor agreement — nothing to enforce.

No owner — it decays like any other unowned system.

No baseline — you cannot show it worked.

Each of these is a management problem. Buying software to solve a management problem produces a licence fee and the original problem.

Signs you are ready to buy

Concrete indicators, rather than a site count.

The asset register exists and is current.

Work orders are being recorded with asset tags, consistently, for at least a quarter.

A preventive schedule is running and someone is checking completion.

Vendor agreements exist with defined rates and response times.

Someone owns the function and has time for it.

The spreadsheet is genuinely creaking — several people need it simultaneously, mobile access matters, or the volume has passed what one person can maintain.

All six. An operator with five of six should fix the sixth before buying, because software will not supply it and will make its absence more expensive.

The wrong indicator is frustration. A bad month of vendor response feels like a software problem and is almost always a process or agreement problem.