What the Pickers Were Already Doing

In a grocery warehouse somewhere in Western Europe, a system called the picker guide told each worker which station to stand at and how many items to pull. Four researchers, Monika Westphal, Fabian Lorson, Andreas Fügener and Alexander Hübner, changed one thing for 42 pickers there. Instead of receiving a task from the machine, a picker could choose the size of the next one, from five options running from 25 items to 105, at the station they had been sent to. The expectation was that a little more say over the work would lift both performance and job satisfaction.¹

Performance went up. Picks per hour in that warehouse rose 9.7 percent between the last week before the change and the last week after it, and the authors' more careful comparison against a second warehouse put the gain at about 20 picks per hour. Job satisfaction went the other way, falling from 3.88 to 2.83 on a five point scale, a drop the authors' own significance figures describe inconsistently. Workers' sense of self determination fell more clearly.¹

What the interviews turned up

The explanation surfaced only after the experiment, in interviews with 31 of the pickers. Long before the researchers arrived, the floor had been running its own system on top of the official one. Pickers had "routinely bypassed the system by manually logging out and switching workstations." Pickers and shift leaders took turns on the unpopular stations, colleagues swapped when one person seemed to be hogging the good ones, and shift leaders made sure the two slow stations were covered.¹ The new arrangement offered a choice management had designed, and in doing so it cut across the one the workers had built for themselves. A follow up experiment, run with 100 paid online participants reading a scenario instead of with the warehouse staff, pointed the same way: the change most likely undermined self determination by disrupting that earlier job crafting, the authors' term for workers reshaping their own jobs.¹

The change handed workers a form of control that management had designed, and in the process it displaced the form of control they had built for themselves.

My read is that a pilot judged on the number it was built to move would have called this a success, and the cost would have shown up somewhere nobody was measuring. The authors put it in their own words: "When formal systems override tacit team norms or remove opportunities for control, job satisfaction may decline even if performance improves (in the short term)."¹ The study has clear limits. It covers one warehouse and 42 pickers, the authors chose which site got the change instead of randomizing it, and eight weeks was too short, they write, to tell whether lower satisfaction would eventually drag performance down. It is also a working paper, dated July 2025, that has not yet been published in a journal.¹

Choice works differently depending on what it replaces

A second field experiment shows how much depends on the setting. Daniel Kwasnitschka and colleagues let workers at two automotive parts plants, one in Germany and one in Italy, choose which machine stoppages to handle when a smartwatch alerted them, across 31,429 tasks over 13 weeks.² Given a choice, workers took about 80 seconds longer to respond and finished the work about 26 seconds faster, and the authors found no significant change in overall productivity. Satisfaction with how tasks were handed out went up, though the authors treat that single survey item as anecdotal.²

So choice lifted satisfaction in one study and lowered it in the other. The plants and the warehouse differ in too many ways for a clean comparison, but the Kwasnitschka paper itself offers a reason the outcomes could diverge: informal arrangements already in place, it notes, can be contradicted by a new choice regime.² What decides the result is what the workers had already built when the new design arrived.

Before you change who assigns the work

I made the case in an earlier piece on a petrochemical plant that the gap between the written procedure and the work as it happens shows up in no report a senior leader receives. The warehouse study adds a cost to that gap. It is also where people keep their sense of control, and a redesign can remove it without anyone intending to.

So the practical order is to find the workarounds before changing the system, not after, the way the researchers had to. Ask people on the floor, in a setting where the answer carries no penalty, how they actually get through a shift. For each answer, ask what the workaround protects, whether the new design keeps it, and who would notice if it disappeared. When a pilot reports, ask for a satisfaction measure next to the performance number, because in this warehouse the two moved in opposite directions.

What are your people doing right now to take some control back from the systems you gave them, and who in your chain of command could tell you?