This time, a small Bayer team took 90 days. Speed depends on where the decision lands.
Bayer’s own example: a small team simplified multi-vial self-infusion for haemophilia patients in one 90-day cycle, instead of the usual two years. Another team moved cancer-medicine production to a new site in about nine months instead of two or three years.
Before, 12 levels of bosses sat between teams and patients. Bill Anderson took over in June 2023, when the 160-year-old company employed 102,048 people, and wrote that its staff were “trapped in 12 levels of hierarchy”.
So in January 2024 Bayer launched 90-day cycles in place of annual planning and moved decisions to the teams closest to the work. It cut half its management layers and 12,000 jobs: by June 2025 the headcount was 89,556, and the cuts fell mainly on management and coordination roles. The thinking draws on Gary Hamel, who argues that layers of management cost more than they control.
Bayer’s share price proves little: it followed the glyphosate weedkiller lawsuits more closely than the organisation chart.
But removing a layer moves a decision. It does not choose where.
Let go of a party balloon, and only the tallest uncle can reach it.
In 2012, Julie Wulf of Harvard Business School published a study of more than 300 large American firms over about 15 years as they removed layers. Decisions drifted upwards, and chief executives grew more involved with their direct reports. Their own reason for flattening: to “get closer to the businesses”. It ended up looking like centralisation at the top.
A pharmaceutical distributor with four signatures between a pharmacist’s order and a delivery van faces the same question. Remove one, and where the decision lands decides whether anything got faster. Before you remove one, name who decides instead.
Speed depends on where the decision lands.
When a manager’s role goes, who decides: the team, or the boss above?
