A fair score for every picker, from a time and motion study

The situation

I was Duty Shift Manager at XPO on the ASOS contract, in charge of a despatch shift at the fully automated fulfilment centre: 26 team managers and, through them, 400 to 800 operatives. Pickers were judged on a raw number — units per hour — and nothing else. That number didn't know whether someone had walked the length of the building for every pick or stood in one bay all shift, or whether their picks were awkward or easy.

So the "best" pickers were often just the luckiest, the others had no reason to try harder, and team leaders had no fair way to manage performance. There was no labour management tool to fix it.

What I did

I built one myself, in Excel, starting with a time and motion study.

1. 01Time itA time and motion study of the pick operation: how long each pick should take, and why.
2. 02Set a fair standardEvery pick weighted for distance walked and how hard it was, so no one was penalised for a bad aisle.
3. 03Score and rankEach picker's performance against standard, shown on leaderboards the teams could see.
4. 04Prove itPiloted on one shift, refined with the operatives' feedback, then rolled out to every shift.
5. 05Hand it overXPO built the model into an industrial timed-performance platform inside the OMS, and used it to pay the pick bonus.

• A handicapped score, not a raw count. The model split productive from unproductive time and weighted every pick for the distance walked and how difficult it was. Each picker got their own tailored target, so the score measured effort and skill rather than luck.
• Leaderboards that meant something. Because the scores were fair, ranking people on them motivated people instead of putting them off. Team leaders could finally have straight conversations about performance, backed by numbers everyone trusted.
• Pilot first. I tested it on one shift, listened to the operatives, adjusted it, then rolled it out across every shift.

What changed

• Faira performance score for every picker, adjusted for distance and difficulty
• All shiftsrolled out after a one-shift pilot
• Paythe model became the basis for the pick bonus

KPIs improved consistently across the shifts. It worked well enough that XPO built my model into an industrial timed-performance platform inside the existing order management system, and used it to pay pick bonuses to staff. A spreadsheet I'd built on a shift became the way the site measured and rewarded its pickers.

What I'd do next

Today I'd build the same thing in days, not months. I'd feed the standards straight from the WMS pick data, and use AI to flag outliers — a location that's always slow, a picker who's struggling — so team leaders spend their time coaching rather than chasing numbers. The principle wouldn't change: measure people fairly and they'll respond.