Calder Distribution: Explaining a Drop in Pick Rate in 90 Seconds
You supervise the outbound floor at Calder Distribution. At Monday's 7 a.m. huddle, in front of eight colleagues, the operations manager asks why the pick rate fell from 92 to 78 units per hour last week. You have four facts: six of the 22 pickers are new hires with three days of training; fast-moving SKUs were relocated to aisle 12, adding roughly 40 metres to each round trip; one handheld scanner failed on Wednesday and was offline for five hours; and overtime was cut from 60 to 20 hours. You have 90 seconds, no slides, and the ops manager follows up with: "So who dropped the ball on training?" Two of the new hires are standing in the room.
1. Give your 90-second response. Write what you would actually say.
Read the worked answer
Model answer (about 85 seconds):
"Short version: 78 is four causes stacking, not one bad week.
First, throughput. Moving the fast-movers to aisle 12 added about 40 metres per round trip. On a full shift that is the single biggest factor and it is a layout decision, not an effort problem.
Second, capacity. We lost the scanner for five hours on Wednesday. That is roughly a third of one shift's picking time gone in one day.
Third, mix. Six of 22 pickers are three days in. At three days, a picker is accurate but slow-that is what the ramp curve looks like, and I would rather have accurate and slow than fast and wrong.
Fourth, hours. Overtime came down from 60 to 20, so we were picking with less coverage at peak.
What I want this week: re-time the aisle 12 route with two experienced pickers so we have a real number instead of my estimate, a spare scanner staged on the floor, and a buddy pairing for the new hires. I will bring you the revised rate Friday."
Why it lands: headline first, causes ranked by size, one ask per cause, and a date. No adjectives about how hard anyone worked.
2. How do you answer "So who dropped the ball on training?" with two new hires in the room?
Read the worked answer
What I would say: "That one is mine to answer-the schedule was set before the six start dates were confirmed, and I approved it. What I would rather talk about is the fix: I want each new hire buddied with an experienced picker for their first four shifts, and I need two people released from picking to make that work. Can we agree that this morning, or do you want it in writing first?"
The reasoning:
- Take the ownership question off the floor immediately. Deflecting ("it's HR's schedule") or distributing blame ("the agency sent them late") reads as evasion in front of eight people and makes the room cautious about raising problems later.
- Do not accept a framing that names individuals. The question asks "who", but the useful answer is a process answer. "I approved it" converts a blame question into a decision question.
- Do not use the two new hires as evidence. "They're doing fine" puts them under scrutiny and sounds defensive; it also invites the manager to assess them in public.
- Convert to an ask with a cost. "Two people released" makes the fix concrete and forces a resourcing decision, which is the substantive issue hiding inside the blame question.
3. What would you deliberately leave out of the 90-second answer, and why?
Read the worked answer
Left out, and the reason:
The estimate that aisle 12 might account for "half" the drop. I do not have the re-timed figure yet. A number I cannot defend becomes the number that gets quoted back at me on Friday. I said "biggest factor" and "my estimate" instead, and committed to a real measurement.
Morale and fatigue. True and relevant over months, but in a 90-second operational answer it reads as an excuse and invites "so what are you doing about it?"-a second conversation I am not prepared for yet.
The scanner's maintenance history. The vendor's service record is a procurement issue; raising it now splits the manager's attention and sounds like blaming a supplier for a throughput decision I own.
Any individual's name in connection with a number. Once a name is attached to a figure in a huddle, the figure stops being a process problem.
Principle: in a short spoken answer, every extra cause dilutes the ones you can act on. Four ranked causes with four asks is already at the limit of what a listener retains in 90 seconds.