In early 2020, distributed work was a fringe benefit at most tech companies — something offered to a small cohort of senior engineers and the occasional remote-friendly startup. By 2021 it had become the industry's dominant operational mode. And by 2024, the backlash was in full swing: high-profile return-to-office mandates from major employers, sweeping headlines about the death of remote work, and a wave of commentary from executives suggesting that in-person collaboration was irreplaceable and that distributed teams had failed the productivity test.
None of that narrative is especially accurate. And the more you dig into what has actually happened to the tech industry's working patterns over the past several years, the more interesting — and less resolved — the picture becomes.
The RTO Narrative Is Overblown
When major tech employers began announcing return-to-office requirements, the coverage treated it as a decisive verdict. But what many of those announcements actually required was two or three days per week in office for employees who lived within commuting distance — which, for many distributed teams, meant nothing changed. The headlines counted these as RTO events. The workers involved continued working from home four or five days a week.
More meaningful is the shift in how companies think about geography when they hire. Even organizations that have reinstated office attendance requirements have typically maintained significantly expanded geographic hiring footprints compared to their pre-pandemic practices. The combination of more flexible policies, better tooling, and a wider talent pool that expects some remote optionality has made fully rigid, office-only hiring a competitive disadvantage in many technical roles.
The companies most aggressively pushing for full in-person attendance tend to share a few characteristics: they're large enough to have considerable negotiating power over workers, they have offices in cities with large local technical talent pools, and they're making bets that the productivity and collaboration benefits of co-location outweigh the talent access costs. Whether those bets pay off will become clearer as hiring market conditions shift over the next few years.
What Actually Changed: Hiring Geography
The most durable shift from the past five years may be the geographic redistribution of tech employment. Prior to 2020, the gravitational pull of a handful of metros — the San Francisco Bay Area, Seattle, New York, Toronto, London — was nearly absolute for engineers seeking roles at well-capitalized tech employers. The implicit rule was that if you wanted to work at a company doing interesting things at scale, you moved to where the company was.
That rule has meaningfully loosened. It hasn't disappeared — clusters still matter for networking, early-career development, and the kind of spontaneous collaboration that's genuinely hard to replicate asynchronously. But the radius around those clusters has expanded, and entirely new concentrations of technical talent have emerged in cities and regions that previously had limited presence in the industry's hiring map.
This has compounding effects. Engineers who can work from smaller cities and regions can afford to be more selective about who they work for, because their cost of living is lower and their leverage relative to local housing costs is higher. Companies that hire broadly benefit from a larger candidate pool, but face new challenges around culture coherence, timezone coordination, and the logistics of keeping distributed teams aligned.
Asynchronous Communication Has Matured
One of the more practical developments of the distributed work era is the improvement — and, more importantly, the normalization — of asynchronous communication practices. Before 2020, the dominant model of workplace communication in tech was heavily synchronous: meetings, Slack responses expected within minutes, real-time standups, and the ambient social pressure of an open-plan office that made extended focused work difficult.
The forced shift to remote work broke some of those habits, though not all of them. Many organizations simply replicated the synchronous model over video conferencing — trading open-plan office interruptions for a continuous stream of Zoom calls. Those organizations, predictably, found remote work exhausting and relatively unproductive.
The ones that adapted more successfully developed genuinely async-first practices: detailed written documentation, recorded video updates that could be watched at variable speed, clear norms around response latency, and meeting-light cultures where synchronous time was reserved for decisions that genuinely required it. Those practices have survived in organizations that have since returned to hybrid arrangements, because the benefits — better documentation, more considered communication, more protected time for focused work — don't disappear when people come back to the office.
The Management Layer Has Been Stress-Tested
Distributed and hybrid work has exposed something many organizations already knew but preferred not to address: that a significant portion of middle management in tech companies was built around physical proximity rather than genuine value-add. Managers whose effectiveness depended on being able to see whether engineers were at their desks, who facilitated unnecessary meetings as a visible demonstration of their own relevance, or who functioned primarily as information relays rather than strategic contributors tended to struggle in distributed environments.
This has pushed more of the industry toward outcome-based management frameworks, where what matters is whether objectives are being met and quality is being maintained — not whether the work is visible in real time. The shift is incomplete and uneven. But it's real, and it's one of the more significant organizational changes that remote work has catalyzed.
The inverse problem has also emerged: some engineers, especially earlier in their careers, have found remote environments genuinely difficult for reasons that have less to do with productivity and more to do with professional development. Learning technical craft through osmosis — overhearing how senior engineers approach problems, asking quick questions, observing how experienced people navigate organizational dynamics — is harder to replicate over Slack and video calls. Organizations that dismissed this concern as executive preference for visibility missed the genuine developmental argument for some structured in-person time.
Tools and Infrastructure Have Caught Up
In 2020, the tooling available for distributed technical teams was functional but rough. Video conferencing frequently failed at scale, collaborative coding environments were limited, and the friction of working across time zones on shared codebases was non-trivial. Five years of significant investment in this infrastructure has produced a meaningfully better stack.
AI-assisted development environments have changed the leverage equation for individual contributors working asynchronously. When a developer can get a capable coding assistant to help them work through a problem at 11pm in their timezone without needing to synchronously engage a colleague, the productivity arguments for overlapping time zones become less absolute. This doesn't eliminate the value of collaboration — it shifts the mix of what benefits most from real-time interaction versus what can be handled effectively independently.
Documentation tooling, code review workflows, and project management systems have all improved in ways that favor distributed teams. The organizations that invested seriously in these systems during the peak remote period tend to operate more efficiently regardless of where people are physically located — because good async infrastructure benefits in-office and remote workers alike.
Culture Is the Harder Problem
The least resolved challenge of distributed tech work isn't productivity, tooling, or management structure. It's culture. How organizations develop and transmit shared values, norms, and ways of working across a geographically distributed team — without relying on the shared physical context that office environments provide by default — remains genuinely unsolved for most organizations above a certain size.
Startups can do it reasonably well because they're small enough that a small number of people can carry culture effectively through direct interaction, even distributed. At scale, the problem compounds. Culture degrades when it isn't actively transmitted, and active transmission over distributed channels requires deliberate effort and investment that many organizations are reluctant to make consistently.
This is probably the strongest genuine argument for some structured in-person time — not that people can't do their individual work effectively remotely, but that culture requires a medium that pure async communication doesn't fully provide. Annual or quarterly offsites, regional hubs, and structured onboarding periods in person are approaches that various organizations have adopted with varying levels of commitment.
Where Things Actually Stand
The tech industry's working arrangements in 2026 are neither the triumphant full-remote future that some predicted in 2021 nor the full return to 2019 norms that some executives would prefer. They're a complicated, employer-specific mix that varies by company size, technical function, market position, and management philosophy.
What's clear is that the structural changes — broader hiring geographies, async communication practices, outcome-focused management, improved distributed tooling — aren't going away. They've been absorbed into how the industry operates, regardless of how any given company frames its office attendance policy. The debate about remote work has been louder than the actual transformation, which, as transformations often do, has been happening more quietly and more permanently than the headlines suggest.