For a century, the office answered a question nobody asked out loud: how do we know work is happening? Presence was the proxy. The building was the management system. Then a global experiment proved most knowledge work travels fine — and the proxy collapsed, leaving companies holding expensive real estate and an unexamined assumption.
The ensuing fight — return-to-office mandates versus remote maximalism — mostly missed the productive question. Not where should people work? but what is an office for, specifically, now?
If the office cannot answer "what is this for?", attendance mandates are just rent justification.
What Offices Are Genuinely Good At
Strip away nostalgia and the honest list is short but real. Trust formation: new relationships — new hires, newly merged teams, repair after conflict — form faster with bodies in a room; video sustains trust far better than it creates it. High-bandwidth problem solving: the messy, interruptive, whiteboard kind of session where ideas need to collide quickly. Apprenticeship by osmosis: juniors absorbing how seniors think by overhearing it — genuinely hard to replicate remotely. Belonging in the flesh: periodic physical gathering does something for commitment that no channel matches.
Notice what's not on the list: routine solo work, video calls, anything done with headphones on. The data from badge readers everywhere confirms what workers already know — much of what offices currently host is work the office makes worse.
Running the Office Like a Tool
1. Program it, don't just open it
A tool has use cases. The office's calendar should reflect them: team anchor days built around collaborative work, onboarding cohorts, quarterly problem-solving intensives, deliberate cross-team collisions. An office you must justify per-use stays sharp; an office that's simply "where you go" dulls into beige.
2. Rebuild the space for its actual jobs
If the office's purpose is collaboration and connection, the sea of desks is inventory error. The conversion that matches reality: fewer workstations, more project rooms, better acoustic variety, genuinely good gathering space — and accept that a smaller, better building used intensely beats a large one used at 30%.
3. Measure outcomes, not occupancy
Badge-swipe targets measure rent amortization, not value. Better questions: do teams that use anchor days well ship better? Does time-to-trust for new hires improve with structured in-person onboarding? Did the quarterly intensive move the problem it gathered for? If presence is worth the commute, it will show up somewhere other than the lobby turnstile data.
The Default Is the Enemy
Both camps in the location war share a mistake: treating one mode as the default that needs no justification. The post-proxy organization justifies both — remote earns its place through documented, asynchronous excellence; the office earns its place by doing the specific things distance does badly. Tools, both. Defaults, neither. The companies that internalize that sentence will spend less on real estate and get more from it than the ones still fighting about days per week.