Maintenance software for a team of five

A five-person maintenance team can run good software. What it cannot do is run a project. Nobody has a spare half day, there is no administrator, and the person who set it up is also the person on call. Every choice should follow from that.

Insist on three things

• •Live in a week, with your own buildings and your own first round, not a sample
• •Changes you can make yourself, without a ticket and without a consultant
• •Requests from people who do not have accounts, so reporting a fault costs nobody a licence

Leave these switched off at the start

Every one of these is useful eventually and poison on day one, because each adds a field somebody has to fill in before the thing they wanted to do.

• •Parts and stores, unless you are already losing jobs waiting for parts
• •Labour time capture, unless somebody bills for it
• •Approval chains, until you know which jobs genuinely need one
• •Custom fields, all of them, for at least two months

Start with the round that scares you

Pick the inspection you would least like to be asked about, put that one on the phone, and run it for a month. It is usually the one with the most steps, which means it benefits most from being guided, and it produces the evidence you were most exposed on.

The one number to watch

Not adoption. Count the proportion of finished jobs that arrive with a photograph. If that number is climbing, the software is being used properly and everything else follows. If it is flat at nothing, people are closing jobs from memory and the rest of the record is decoration.

What small teams get that large ones do not

You can change your mind in an afternoon. Use it. Run something for a month, keep what worked, delete the rest. Large organisations pay consultants for the privilege of doing badly what you can do over a coffee.