The "return to office" debate of 2023–2025 ended in a quieter conclusion than anyone predicted: most engineering work didn't return. Some companies forced compliance for cultural or political reasons. Most discovered that the work they hired engineers to do had never genuinely required colocation. The 9-5 office model, built for industries where presence matters, doesn't fit a category whose primary output is text — code, documentation, design files, written decisions.
This piece walks through five structural reasons the 9-5 colocated office model is dead for engineering work, what's replacing it, and why the VDC model is one of the natural successors.
Reason 1: The work product doesn't require physical presence
An engineer's primary output is text. Code in repositories. PR descriptions. Design docs. Architecture diagrams in collaborative tools. Test results. Review comments. None of this requires anyone to be in the same building. The artifacts move through cloud-hosted version control, asynchronous review tools, and collaborative documents — all designed to function regardless of physical location.
The work that does benefit from physical presence — visual whiteboarding, intense pair-debugging, social rapport — accounts for 5–15% of an engineer's week. Designing a colocated office around the 5–15% case while the 85% case is asynchronous is structurally inverted.
Reason 2: Talent geography unbundled from work geography
The 9-5 office assumed your talent pool was the local labor market. In 2026, this assumption is empirically wrong. Senior engineers in any given city compete with senior engineers globally for the same roles. Compensation has converged across major tech hubs (the New York / SF premium is smaller than it was; the Bangalore / Lisbon / Mexico City discount is smaller than it was). The frictions that kept geography aligned with work — visas, time zones, coordination cost — are structurally lower than they were.
An engineer in Lisbon with the right skills can work for a US-based company with the same effective overhead as a Boston-based engineer. The Boston engineer's geographic premium is no longer compensated by structural advantages.
Reason 3: The async-first stack works
The tooling that makes async-first engineering practical hit production maturity around 2018–2022. GitHub. Linear. Notion. Slack. Loom. Async-first design tools (Figma, Miro). Each individual tool replaced an in-person ceremony with an async equivalent that worked at least as well, often better.
By 2026, async-first isn't a workaround for remote teams; it's a productivity choice that even colocated teams adopt because it produces better artifacts (searchable, durable, asynchronously consumable). When the tools are async-first, the question of where engineers physically are becomes operational noise.
Reason 4: Engineering productivity research reframed the question
Research from 2020–2024 (DORA, SPACE framework, productivity studies from Microsoft and others) consistently showed:
- Productivity correlates with focus time, not hours-in-office.
- Remote engineers report more focus time than colocated ones.
- The collaboration that benefits from colocation can be replicated through deliberate sync windows.
The 9-5 office optimizes for "presence" as a proxy for productivity. The proxy was always weak; the data killed it. Boards and CFOs who track engineering output (rather than engineering attendance) have shifted accordingly.
Reason 5: The cost structure favors distribution
Real estate. Office facilities. Commute time. Local labor markets. Each of these is a cost the 9-5 office implicitly imposes — on the company (rent, infrastructure) and on the employee (commute, geographic constraint, lifestyle inflexibility). Distributed work removes most of these costs without proportional value loss.
Companies that ran controlled experiments (some teams 9-5 colocated, others distributed) consistently found similar engineering output with significantly lower facility cost. The math doesn't justify the infrastructure.
What's replacing the 9-5 office
Three patterns dominate:
- Distributed-first with deliberate sync. Engineers anywhere, with structured overlap windows for synchronous work. Sprint cadence, retrospectives, occasional in-person meetups.
- Hub-and-spoke. A core in-person hub for a fraction of the team plus a distributed network of remote engineers. Hybrid that respects both modalities.
- Pod-based delivery (VDC pattern). Cross-organizational pods that operate distributed by default, with embedded delivery management replacing the colocated-EM function.
Each pattern has different operational characteristics, but all share the rejection of the 9-5 colocated assumption.
Where the 9-5 office still makes sense
To stay honest:
- Hardware engineering. Physical lab work, prototype assembly, electronics testing — these benefit from colocation.
- Heavy customer-facing work. Sales engineering, deeply customer-embedded consulting where in-person presence at the customer matters.
- Specific cultural/onboarding contexts. Brand-new engineers, especially early-career, often benefit from in-person mentorship that's hard to replicate remotely.
Software-engineering work that's primarily codebase-shaped doesn't fit any of these. The 9-5 office's place is narrower than it used to be — but it isn't zero.
What this means for engineering leaders
Three implications:
- The talent pool is global. Recruiting that constrains to local markets is leaving 90% of qualified candidates on the table.
- The cost structure is reshapable. Distributed teams produce output at lower facility cost without proportional productivity loss.
- The engagement models scale. VDC pods, BOT engagements, hybrid in-house+distributed structures — all become viable when colocation isn't the constraint.
Frequently asked questions
What about culture and team cohesion?
Culture exists in distributed teams; it just looks different. Slack channels, retrospectives, occasional in-person meetups, shared values documented and lived rather than assumed. Companies that built distributed culture deliberately (GitLab, Automattic, others) have published their playbooks. It's possible; it requires intent.
Don't junior engineers benefit from in-person mentorship?
Some do. Distributed-first companies often invest in periodic in-person concentrations (week-long meetups, quarterly offsites) to address this. The 9-5 office isn't the only way to provide mentorship.
Will the office come back?
For software engineering specifically, no — at least not as the default. Some companies will mandate office presence for cultural or political reasons; competitive pressure for talent will push back against this. The trend is one-way.
What's the engineering manager's role in distributed teams?
Different but not diminished. EMs in distributed teams focus on async-first leadership, structured sync time, async decision-making, and explicit culture-building. The skill set evolves; the role doesn't disappear.
Where to start
If your engineering organization is debating return-to-office or hybrid-vs-distributed, the data favors distributed by a wide margin. The harder questions are operational: which teams need physical concentration (hardware? early-career-heavy?), and which teams should be distributed by default (most software engineering).
For broader workforce-strategy conversations, schedule a 30-minute call. For distributed-pod operating patterns, see communication patterns for remote VDCs.