The short answer
The user is standing up, often wearing gloves, in a building with patchy signal, on a device shared across a shift. Every design decision follows from that.
If a task takes more than a few taps, it will be batched at the end of the shift from memory, which is worse than not recording it.
What the context demands
- Large tap targets, usable with gloves
- Very few taps per task, ideally one or two
- Full offline working with visible sync status
- Fast sign-in on a shared device, and automatic sign-out
- Readable in poor light and in bright light
- Minimal typing, with selection instead wherever possible
The typing point is the one most often missed. A free-text field in a corridor gets one word or nothing.
Reduce what has to be captured
| Instead of asking | Derive it |
|---|---|
| Which site | From the location or the schedule |
| What time | From the device |
| Which task | From the schedule for that slot |
| Who | From the signed-in user |
| How long | From start and finish taps |
Once those are derived, the only thing a person has to record is the exception, which is a much smaller ask and much more likely to happen.
Make exceptions easy
Reporting a problem should be faster than ignoring it. A photo, a category and a tap is about the limit of what will be done reliably in the middle of a shift.
And something has to visibly happen afterwards. Staff who report a fault and see nothing change stop reporting, and then you have lost the channel.
Language and literacy
Cleaning teams are frequently multilingual and some staff have limited reading confidence in the language of the app. Icons, colour and short phrases carry more than sentences.
Offering the interface in the languages your teams actually speak is usually a modest piece of work and makes a substantial difference to adoption.