Return-to-Office · Primly Community

remote vs hybrid vs in-office productivity, my actual data after tracking for a year

sre_sol · 4 replies

I started logging this when we got our hybrid mandate, because I was tired of arguing about it based on vibes. 12 months of notes. Take this as one person's data point, not a study.

Context: SRE, senior. Work is half async coordination (incidents, runbooks, infra work), half synchronous (incident response, planning, cross-team debugging).

Fully remote days (home): deep focus work is better. Runbook rewrites, infra migrations, documentation. Zero interruptions. Average focused session: 3.2 hours before a break. Also: on-call response time faster because I'm already at my desk, no commute overhead.

In-office days (3 days/week): sidebar conversations are genuinely useful. Caught two architectural issues in hallway chats that I probably would have missed. Meeting energy is different face-to-face. But: open floor plan, so 60% of the time I'm wearing noise-canceling headphones and pretending I'm at home anyway.

Surprise finding: my worst productivity days were forced in-office days that I spent entirely on Zoom. I drove 45 minutes to sit on the same video calls I could have done from home. That's not an opinion, that's just waste.

What actually helps regardless of location: team agreements on async/sync norms, documented decisions, response-time expectations. The teams I work best with have this regardless of where they sit. The teams that are chaotic are chaotic in the office too.

I'm not anti-office. I'm anti-office-as-performance. My company's RTO justification was "collaboration and culture." But we never defined what collaboration looked like or measured it. We just counted badge swipes.

Badge swipe rates are not a productivity metric. That's the part I can't get past.

4 replies

mobile_mara

"badge swipes are not a productivity metric" should be embroidered on a pillow and placed in every exec conference room. we track lines of code, feature velocity, incident rate. nobody tracks whether the in-office days specifically moved those numbers.

analyst_ana

As someone doing data work: have you thought about running a proper comparison, like output metrics on office days vs remote days? I know individual work makes confounding hard but even a rough look would be interesting.

sre_sol

I tried. The problem is my output isn't consistent week to week based on incident load, which is randomish. So it's hard to isolate location as the variable. I looked at PR review turnaround times though, and those were slightly faster on remote days. But n=1 and I wouldn't make a policy argument from it.

brand_ben

As a manager I'll add: I've managed fully remote teams that ran circles around co-located teams and vice versa. The location isn't the variable most people think it is. Clarity of goals and psychological safety have more predictive power in my experience.