Ridgeway Device and three different answers

We run three device counters in the same city. Screens, batteries, water damage, the occasional laptop board. High volume, low ticket, and a promise of same-day on most of it — which means the whole business runs on whether we can tell what is happening right now.

For about two years I could not. Not because nobody was measuring, but because all three branches were.

Three branches, three definitions

The east branch counted turnaround from when the customer handed the device over. The north branch counted from when a technician picked it up, which on a busy Saturday could be four hours later. The third counted from the ticket being written, which happened whenever somebody got round to it.

So every month I got three numbers that all looked reasonable and could not be put next to each other. When the east branch reported a bad week I could not tell whether it had been a bad week or a strict definition.

Worse, none of them could hand a job to another branch. A customer who dropped a phone at one counter and wanted it from another was a phone call between two managers and a note on a bag.

One definition, imposed

The reason we moved was not the reporting. It was that I wanted one intake form that all three branches had to use, so the argument about definitions would stop being an argument.

Intake now takes the IMEI, the serial, the passcode handling, and a photographed condition check before the device leaves the counter. That last one settled more disputes in the first month than anything else we have ever done. What the customer agreed to on data — whether a backup was taken, whether the device gets wiped — is on the record rather than in somebody's memory.

Because it is one system, a job taken in at one branch is visible at all three. The transfer conversation stopped happening.

What changed

The number I care about is that turnaround is now one number. Median across all three branches, measured from the same event, which is the moment the device is booked in. It is a little under two days across everything and about four hours on a screen, and when a branch drifts I can see it in the week rather than in the month.

Volume is up roughly a quarter on the same three counters and the same staff. Most of that is not efficiency in the workshop; it is that the counter is not repeating itself. The messages go out when the device is on the bench and when it is done, and the same-day customers stopped ringing at four o'clock to check.

The part I would not give up is the quote before we open anything. It is agreed on the customer's own page, with the date it was agreed, and it has taken a whole category of conversation out of the shop.