Uber is cutting roughly 3,300 corporate roles — about 10% of its non-driver workforce, and its largest reduction since 2020. That is the number the coverage led with. It is not the number that explains the decision.
Inside the same announcement is a structural figure: Uber has reduced the population of employees sitting seven or more layers below the chief executive by 20%, and cut the number of “micro-teams” — managers with only one or two direct reports — by nearly half. CEO Dara Khosrowshahi framed the reorganisation as consolidating places where, in his words, fragmentation was creating duplication and slowing decisions.
Read the first half of that sentence again. Uber had enough people standing eight, nine and ten rungs down from the chief executive that it needed a formal programme to count them and a target to shrink them.
Headcount is the symptom. Depth is the disease.
Headcount tells you what a company costs. Layer depth tells you how long it takes to say yes. Every additional layer between the person who spots something and the person who can fund it adds a review, a deck, a calendar constraint and a translation step — and each translation shaves off a little more of the original signal. Companies rarely decide to become seven layers deep. They arrive there one well-intentioned promotion at a time.
The micro-team figure is the more damning of the two. A manager with a single direct report is not managing anything; they are holding a title that was almost certainly created to solve a compensation or retention problem for one individual. It is a cheap fix at the moment it happens. The cost shows up later, permanently, as an extra layer sitting on top of everyone underneath.
Uber is also concentrating teams into a smaller set of hubs — New York and San Francisco chief among them — and asking most remote staff to relocate, taking the fully remote share down toward roughly 1%. Whatever you think of return-to-office as policy, geography and layer count are the same lever pulled twice: both are attempts to shorten the distance between a question and an answer.
Our take: Nobody outside Uber should care how many people Uber employs. Everybody running an organisation should care that a company this sophisticated let ten layers accumulate without noticing. Layer depth is not tracked on any standard dashboard, which is exactly why it drifts. It is also the cheapest thing on this list to measure: you can count it this afternoon from your own org chart, and you do not need a restructuring to fix a manager with one report.
What to watch
- Whether the layers stay gone. Depth is easy to cut once and easy to rebuild quietly over four promotion cycles. The tell is whether Uber reports this metric again in twelve months or lets it disappear.
- Severance and charges in the next quarterly filing. A 3,300-person reduction carries a one-time cost that will land before any of the savings do.
- Attrition in the hub cities. Relocation mandates tend to produce a second, unbudgeted wave of departures — often from the people with the most options.
- Copycats. “Cut layers, not just people” is a framing other large-cap operators will find attractive precisely because it sounds like discipline rather than distress.
The version of this that applies to you
You do not need 30,000 employees for this to bite. Open your org chart and count two things: the longest path from you to the person furthest from you, and the number of managers with fewer than three direct reports. If the first number is above five in a company under a few hundred people, decisions are taking longer than you think they are. If the second number is above zero, you have at least one layer you created for a reason that had nothing to do with how the work flows.
Uber needed a 3,300-person restructuring to get at it. Most companies can get most of the benefit with a conversation and a redraw.