A worker arrives at the office, swipes the badge, is seen by the right people, drinks a coffee, and goes home to do the actual work. The maneuver has a name now, coffee-badging, and it captures a broader truth about mandates: any rule that counts attendance will be met with the minimum attendance that counts. The employee is not defying the policy. They are complying with it exactly, and in doing so revealing what it was really asking for.
This is malicious compliance in its purest form, and it is worth being precise about why it appears. A mandate that says "be productive" is hard to game, because productivity is the thing being asked for. A mandate that says "be in the building three days a week" is trivial to game, because the building is a proxy for productivity, and proxies can be satisfied without satisfying the thing they stand for. People do not game goals; they game measures. When the measure and the goal come apart, the measure wins, because the measure is what gets checked.
The benefit nobody dares to use
Semco once ran into the same dynamic from the opposite direction. As Semler has told it, in the 1980s the company offered every new father two weeks of paid paternity leave, and not one of them took it, out of fear of what the boss would think. The rule existed; the behavior it invited did not, because a stronger unwritten rule overrode it. A benefit nobody dares to use, Semler concluded, is not a benefit but decoration. Coffee-badging is that story inverted. The written rule demands presence, and the unwritten calculus, get seen, protect yourself, do the real work elsewhere, quietly wins.
Both cases point to the same lesson: formal policy sits on top of a stronger layer of incentives and fear, and when the two conflict, the informal layer decides behavior. A mandate that ignores this does not get compliance. It gets the performance of compliance, which is worse, because it looks like success on the attendance report while hollowing out the thing the report was supposed to protect.
Presence theater and its costs
The deeper cost is that presence theater degrades the very focus the office was meant to provide. Microsoft's telemetry across tens of thousands of workers found employees interrupted roughly every two minutes during the workday, adding up to hundreds of interruptions a day, with after-hours chats rising year over year. Forcing people into that environment and then measuring their attendance produces a specific outcome: they show their faces during the interruptive hours to satisfy the badge, and do their concentrated work at the margins, early, late, or from home. The mandate has not gathered the work; it has scattered it and added a commute.
There is also a trust cost that compounds. When a policy is transparently a proxy, employees learn that the organization values the appearance of a thing over the thing itself. That lesson generalizes. People who conclude that management cannot tell the difference between being seen and doing good work will, reasonably, invest in being seen. The mandate trains the workforce in exactly the behavior it claims to oppose.
Read the gaming as data
The productive response is not tighter enforcement, more turnstiles, minimum hours on site, badge audits, because tighter enforcement of a bad measure just produces more elaborate gaming. The productive response is to treat the coffee-badging as information. If people will do almost anything to be seen without working, and almost anything to work without being seen, the policy has inverted the relationship it wanted. It has made presence the goal and work the workaround.
A leader who notices this has a cleaner path available: measure the work. Ask whether the deliverable shipped, whether the customer was served, whether the team hit what it committed to. Those measures cannot be satisfied by a coffee and a badge swipe. They are harder to check, which is precisely why mandates avoid them, and precisely why they are the only measures that malicious compliance cannot defeat. When the rule finally asks for the real thing, the theater has nothing left to perform.
Sources
- Bloom, Davis & Barrero / Stanford (2025). Return to office? Not everybody is doing it. In late-2024 survey data, only 44% of US workers said they would comply with a fully in-person RTO policy; the rest would quit or job-search. https://news.stanford.edu/stories/2025/03/return-to-office-not-everybody-is-doing-it
- Microsoft WorkLab (2025). Work Trend Index: The Year the Frontier Firm Is Born. Microsoft 365 telemetry: workers interrupted roughly every 2 minutes, ~275 interruptions/day, after-hours chats up 15% year over year. https://www.microsoft.com/en-us/worklab/work-trend-index/2025-the-year-the-frontier-firm-is-born