By mid-2026, a small number of the largest, most visible employers (Amazon and JPMorgan Chase among them) have moved from hybrid schedules to full five-day, no-exceptions office mandates. JLL’s own tracking, reported directly by Bisnow, found that more than half of Fortune 100 employees are now required to be in the office five days a week, up from just 11% a year earlier.
Yet the share of the US workforce actually working from home has barely moved: Stanford’s own Survey of Business Uncertainty research, run by Nicholas Bloom and Steven Davis with the Federal Reserve Bank of Atlanta, found that work-from-home accounts for roughly 21.2% of all paid workdays and projected that even every currently planned RTO mandate would only trim that to about 20.8%.
That gap between headline mandates and aggregate behavior is not a paradox. It reflects the fact that these mandates are concentrated among a small number of very large, very visible employers, while most of the broader market, including the majority of smaller and mid-sized companies, never adopted a strict five-day policy in the first place.
A second, separate gap matters just as much for engineering leaders: most return-to-office research measures attendance and self-reported satisfaction, not code velocity, deployment frequency, or incident response time.
If you lead a distributed engineering team, the return-to-office debate is not really about whether offices are useful, but whether mandating office time improves the specific outcomes you are accountable for, and whether the retention and hiring costs of a return-to-office mandate are worth paying.
This article breaks down what the 2026 data actually shows for engineering teams, where the evidence is strong, where it is thin, and what to do operationally before a policy change costs you your most experienced engineers.
In this article:
What 2026 RTO Data Actually Shows
- How Common Office Mandates Are in 2026
- What the Research Actually Measures
Why Broad Office Data Does Not Map Cleanly to Engineering Teams
How Work Model Changes Affect Engineering Output
- Deep Work, Interruptions, and Calendar Fragmentation
- Async Code Review, On-Call Coverage, and Time Zone Overlap
- When In-Person Collaboration Helps and When It Doesn’t
Retention Risk, Return to Office Morale, and Team Stability
- Who Leaves First After a Return to Office Mandate
- How Enforcement Drives Attrition More Than Policy Language
- What Mandates Change for Work-Life Balance and Employee Engagement
Hiring Implications for Distributed Engineering Orgs
- How Location-Bound Policies Shrink the Engineering Talent Pool
- Time-to-Hire, Backfill Risk, and Recruiter Bandwidth After Attrition
- Where AI-Assisted Hiring Changes the Tradeoffs
Comparing Fully Remote, Hybrid, and Full Office Models
Pros and Cons of Return to Office Mandates for Engineering Teams
How Office Design and Schedule Design Shape Outcomes
A Practical Decision Framework for Engineering Leaders
- Questions to Ask Before Changing Your Work Model
- How to Structure a Hybrid Policy Without Creating Two Teams
- What to Measure After Rollout
Frequently Asked Questions
What 2026 RTO Data Actually Shows
Most of the headline return-to-office statistics you see in 2026 describe office workers in broad terms. They do not isolate engineering teams, and the difference matters.
How Common Office Mandates Are in 2026
Return to office mandates are common, but “common” hides a lot of variation depending on what you’re measuring.
According to KPMG’s 2024 U.S. CEO Outlook, a survey of 1,325 CEOs globally including 400 in the US, fielded in July and August 2024, 79% of US CEOs expected traditionally office-based roles to be fully back in the physical workplace within three years, a sharp jump from 34% earlier that same year. Only 17% envisioned a hybrid setup, and just 4% expected fully remote to persist.
The same survey found 86% of CEOs planning to reward employees who make the effort to come in with better assignments, raises, or promotions. That is executive intent, not universal current policy: as the JLL data above shows, just over half of Fortune 100 companies have actually implemented a five-day mandate so far, and the picture across mid-sized and smaller employers is far more hybrid.
In other words, full five-day mandates are where a growing share of CEOs want to end up, but most companies, including most engineering-heavy ones, have not gotten there yet.
The picture at the very largest employers, described above, is driven by a specific set of sectors, especially finance and e-commerce, going all the way to five days, while most large tech employers have settled on a three-day hybrid floor instead. The list below shows exactly that split.
Specific 2026 policies for large tech and finance employers:
- Amazon: five days in office, in effect since January 2025 for corporate employees
- JPMorgan Chase: five days in office for essentially its entire 300,000-plus global workforce, effective March 2025, ending a policy that previously required this only of managing directors and client-facing staff
- Google: three days in office per week for most hybrid roles
- Meta: three days in office per week
- Apple: three days in office per week (Tuesday, Thursday, plus a team-specific third day), with Apple’s own leadership stating publicly that a full five-day mandate is not planned
- Microsoft: expanded its hybrid requirement to three or more days per week starting in 2026, rolling out first in the Puget Sound region
- Shopify: still fully remote-first domestically as of 2026, with no office attendance requirement for most roles
Despite this, JLL’s own Global Occupancy Planning Benchmark Report 2026, drawn from 84 organizations representing 716 million square feet of office portfolios worldwide, found that actual office utilization reached only 56% in 2026, up from 54% in 2025 and 49% in 2024, but still well short of the pre-pandemic level of 61%.
The same report found that organizations enacting office mandates saw the biggest utilization gains of any single lever, but even then, the gap between actual and target utilization was 18 percentage points, the narrowest since the pandemic, but still a real gap. The mandates are working, to a degree, but full compliance and full office usage are still not the same thing.
What the Research Actually Measures
Gallup’s workplace surveys track engagement, satisfaction, and self-reported productivity. The Conference Board tracks HR leaders’ perceptions of hiring and retention difficulty. Stanford’s WFH Research project and Survey of Business Uncertainty track structured, longitudinal data on where and how people actually work. FlexJobs runs an annual State of the Workforce survey of thousands of US professionals.
None of these directly measure engineering velocity metrics like pull request throughput, deployment frequency, mean time to recovery, or sprint completion rates. They measure things like “do you feel productive” and “are you satisfied with your work arrangement.” Those are useful signals, but they are not the same thing as measuring whether your team ships faster in an office.
A few engineering-relevant data points have emerged from more rigorous, causal research:
- A 2024 study from researchers at the University of Chicago and University of Michigan (Van Dijcke, Gunsilius, and Wright) matched 260 million resumes against employment data at Microsoft, SpaceX, and Apple. It found that after each company’s RTO mandate took effect, the share of senior employees dropped: over 5 percentage points at Microsoft, about 4 points at Apple, and about 15 points at SpaceX, which had the strictest policy of the three. Departing senior staff disproportionately moved to direct competitors.
- Separate research from Baylor University and the University of Pittsburgh, “Return to Office Mandates, Brain Drain and Gender Difference” (Ding, Ma, Xing, Yang, and Jin), tracked over 3 million tech and finance employees across S&P 500 firms and found an average 13 to 14 percentage point rise in abnormal employee turnover following an RTO mandate, with the effect strongest among women, senior employees, and more skilled workers. Baylor’s own summary of the findings adds that job vacancy duration rose 23% and hire rates fell 17% at the same firms.
- University of Pittsburgh professor Mark Ma’s analysis of large-scale Glassdoor review data found that RTO mandates were frequently announced after a company’s stock price had already dropped, and that the mandates themselves produced no measurable lift in firm value, alongside a clear drop in employee satisfaction scores.
- Microsoft and GitHub’s joint Developer Experience Lab research, a study of more than 2,000 developers published as “DevEx in Action,” found that deep, uninterrupted work is the single strongest lever for individual developer outcomes, ahead of tooling, feedback speed, or how engaging the work itself is. Developers with dedicated deep-work time reported feeling 50% more productive than those without it, a larger effect than any other factor the study measured.
Why Broad Office Data Does Not Map Cleanly to Engineering Teams
Software engineering work has structural differences from most office jobs that make general RTO data a poor proxy.
Deep-work dependency
A backend engineer debugging a distributed-systems issue or an infrastructure engineer writing Terraform modules needs two-to-four-hour uninterrupted blocks. Open-plan offices and mandatory in-person meetings fragment those blocks.
The most rigorous evidence for this comes from outside tech entirely: Nicholas Bloom’s randomized controlled experiment at Ctrip, published in the Quarterly Journal of Economics, randomly assigned call-center employees to work from home or in the office for nine months and found a 13% performance gain for home workers, driven partly by a quieter, less interrupted working environment, plus fewer sick days and breaks.
Asynchronous workflows
Code review, CI/CD pipelines, and incident response queues run 24 hours a day on distributed teams. Mandating office presence in a single time zone collapses the coverage window and forces synchronous review bottlenecks.
Hiring pool composition
FlexJobs’ 2025 State of the Workforce data shows most remote-capable workers now draw a hard line on office mandates, with 76% saying they’d look for a new job if remote options disappeared entirely.
For senior engineers with a decade or more of experience, and with the most external options, that resistance is generally higher. The talent pool you can access shrinks the moment you attach a strict location requirement.
When you read a headline like “RTO boosts collaboration,” ask what kind of collaboration, for whom, and whether it was actually measured or simply assumed. For engineering teams, the answer is usually assumed.
Read more: Does Remote Work Increase Productivity? What the Evidence Says and What Actually Drives Output
How Work Model Changes Affect Engineering Output
Engineering productivity is a set of interlocking activities, and each one responds differently to where the work happens.
Deep Work, Interruptions, and Calendar Fragmentation
Remote engineers need longer uninterrupted coding blocks than in-office peers for the same reason the Ctrip research above points to: offices generate ambient interruptions that break focus.
The mechanism is straightforward: offices generate ambient interruptions that break focus, and research on context switching from UC Irvine’s Gloria Mark shows it takes roughly 23 minutes on average to return to full concentration after that.
Hybrid schedules introduce a second problem: calendar fragmentation. When your team is in office on Tuesday and Thursday, meetings cluster on those days. In-office days become meeting days, and remote days become the only time deep work actually happens. If you are not intentional about protecting focus time on in-office days, you end up paying the cost of commuting and the cost of lost deep-work hours at the same time.
The fix is not simply choosing remote versus office. It is designing your calendar and meeting norms around the kind of work that drives output. A meeting-free day in office is more valuable than a remote day with six video calls.
Async Code Review, On-Call Coverage, and Time Zone Overlap
Distributed teams operating across three or more time zones gain a structural advantage in code review and on-call rotation, but only if their async processes are mature.
When a team spans US Pacific, Central European, and Southeast Asian time zones, a pull request opened at 5 PM in San Francisco gets reviewed by 9 AM in Berlin and refined by noon in Ho Chi Minh City. That same PR on a co-located team waits until the next morning, because everyone left the office at the same time.
On-call coverage follows the same logic. A follow-the-sun model means no single engineer carries a 2 AM pager burden every week. Mandating office presence in a single location eliminates that advantage entirely.
The tradeoff is real: time-zone-distributed teams need stronger written communication norms, more detailed PR descriptions, and clearer runbook documentation. These are one-time process investments, not ongoing productivity drains.
Read more: On-Site vs Remote vs Distributed Engineering Teams: How to Choose?
When In-Person Collaboration Helps and When It Doesn’t
In-person time has measurable benefits for specific activities:
- Initial team formation. New teams that spend their first one to two weeks together in person tend to build trust and shared context faster than fully remote onboarding alone.
- Complex architectural decisions. Whiteboard sessions for system design, where the problem space is ambiguous and the group needs to converge on a shared mental model, tend to work better synchronously and in person.
- Conflict resolution. Interpersonal friction is harder to resolve over video. Tone, body language, and informal conversation help de-escalate.
In-person time does not show a measurable improvement in:
- Sprint execution velocity (story points completed, bugs resolved)
- Code quality metrics (defect density, review thoroughness)
- Individual contributor output on well-scoped tasks
The implication: a quarterly or biannual team offsite focused on planning and relationship-building captures most of the real in-person benefit. A five-day office mandate captures little additional benefit on top of that, while adding commute time, real estate cost, and retention risk.
Retention Risk, Return to Office Morale, and Team Stability
Return to office morale drops are not hypothetical, though the exact size of the drop depends on which dataset you trust. What is consistent across independent research is the direction and the mechanism: strict mandates lower satisfaction and raise attrition risk, concentrated among the people you can least afford to lose.
Who Leaves First After a Return to Office Mandate
Gartner’s own research, a survey of 2,080 knowledge workers fielded in May and June 2023, has flagged three groups as most likely to leave after a rigid RTO mandate: high performers, women, and millennials.
Intent to stay dropped 8% among average employees under strict mandates, but 16% among high performers, twice the average rate, and 11% among women and 10% among millennials. For engineering teams, the high-performer finding is the most operationally damaging.
Senior engineers and staff-level ICs have the most external options. They are the most likely to have competing offers or established networks that let them move quickly, and they tend to value autonomy and deep-work time more than junior engineers, which makes office mandates feel like a direct downgrade in working conditions.
This isn’t just stated preference. It shows up in actual departures: the Chicago/Michigan resume study and the Baylor/Pittsburgh research cited above both found senior and higher-skilled employees leaving at disproportionately higher rates within roughly two quarters of a strict mandate, often to direct competitors offering remote or hybrid work. The people you most want to keep are the ones most likely to walk.
How Enforcement Drives Attrition More Than Policy Language
The language of a return to office mandate matters less than how it is enforced. Companies that announce a policy but actively track badge swipes and issue warnings tend to see sharper morale declines than companies with a similar policy but looser enforcement.
Pew Research Center found that 75% of workers were required to be in the office a set number of days per week or month as of October 2024, up sharply from 63% in February 2023, yet 46% of remote-capable workers said they’d be unlikely to stay if remote work were removed entirely. A stated requirement and a workforce that actually complies with it are two different things, and the gap between them is where most mandates in practice live.
This creates a real tension. Mandate without enforcement signals distrust and fails to achieve the stated goal. Mandate and enforce strictly, and you risk losing your most experienced people, as JPMorgan’s very public internal pushback in 2025 illustrated: thousands of employees signed an internal petition citing desk shortages, inadequate infrastructure, and lost flexibility within weeks of the policy taking effect.
“Quiet staying,” or remaining employed but disengaging, is the less visible cost. Engineers who feel forced into an office they did not want tend to reduce discretionary effort: fewer after-hours deploys, less mentoring, fewer contributions to internal tooling. You keep the headcount and lose the output.
What Mandates Change for Work-Life Balance and Employee Engagement
Flexible work options are among the strongest predictors of engineering employee engagement in current workforce data, on par with or ahead of compensation and title for many respondents. Removing that flexibility through a mandate pulls one of the levers that most directly controls engagement.
The impact on work-life balance is concrete: A five-day office mandate in a metro area typically adds 50 to 80 minutes of commute time per day. That is five to seven hours per week lost from personal time, exercise, family, or sleep, a cost that falls disproportionately on working parents and caregivers.
Hiring Implications for Distributed Engineering Orgs
A return-to-office mandate is also a hiring policy. The moment you require on-site presence, you change who can apply, how fast you can fill roles, and what it costs to backfill attrition-driven vacancies.
How Location-Bound Policies Shrink the Engineering Talent Pool
If your office is in San Francisco and you mandate five days on-site, your candidate pool is limited to engineers who already live in the Bay Area or are willing to relocate. That is a small fraction of the global engineering talent supply.
The numbers are stark. Arc’s own network includes more than 450,000 remote-ready professionals across 190 countries. A location-bound posting in a single metro area typically draws from a candidate pool that is many times smaller, depending on the role’s specialization.
For niche skills such as Rust systems engineering, ML infrastructure, or platform reliability, the local pool at any given time may have only a handful of qualified candidates.
Remote and hybrid postings widen the candidate funnel significantly. Peer-reviewed research led by Zhenyu Liao, published in Administrative Science Quarterly and based on more than 50 million job postings across 28 European countries, found that remote postings attracted anywhere from 20% more to 10 to 20 times more applicants than otherwise similar in-person roles, giving employers a large enough pool that many raised their skill requirements just to manage the volume.
Time-to-Hire, Backfill Risk, and Recruiter Bandwidth After Attrition
When senior engineers leave after an RTO mandate, backfilling those roles takes longer than hiring into net-new positions. The reasons compound:
Institutional knowledge loss
A staff engineer who leaves after four years takes domain context, system knowledge, and tribal process understanding with them. The replacement typically needs three to six months of ramp-up time, even if they are technically strong.
Recruiter bandwidth
If a mandate triggers a wave of attrition, even five to ten departures in a 100-person engineering org, your recruiting team is suddenly running multiple senior-level searches at once. A search for a senior IC or staff engineer commonly takes 45 to 90 days. Running five simultaneously with the same recruiter headcount means either slower fills or a lower hiring bar.
Candidate hesitation
Engineers evaluating an open role will ask why it’s open. “The previous engineer left after our RTO mandate” is not a compelling answer, and it signals to strong candidates that more departures may follow.
SHRM (the Society for Human Resource Management) estimates that replacing an employee costs 50% to 200% of their annual salary depending on seniority and specialization, and senior, specialized technical roles, like a senior software engineer, sit at the higher end of that range once you count recruiter time, interview hours, signing bonuses, and reduced output during ramp-up.
Where AI-Assisted Hiring Changes the Tradeoffs
AI-powered hiring tools have shifted the math on how fast you can source and vet distributed engineering candidates. If your concern about remote teams is that you cannot find and screen good people quickly enough, that concern has weakened considerably since AI-assisted sourcing and screening tools matured.
Arc’s own AI recruiter, HireAI, is a direct example of the pattern: it analyzes a job’s requirements against a pool of pre-vetted engineers and returns a shortlist of matched, interview-ready candidates almost instantly, compressing a sourcing phase that used to take weeks into a single day.
Structured remote interviews combined with async coding assessments now produce hiring signals that are at least as predictive as traditional on-site interview loops, based on comparisons from companies running both formats side by side.
The operational implication is that if you can hire distributed engineers faster, with comparable quality signals, the argument that “we need everyone in the office to maintain team quality” loses its foundation. You do not need geographic proximity to vet talent. You need good tooling and a disciplined process.
Comparing Fully Remote, Hybrid, and Full Office Models
The right work model for your engineering team depends on what you are optimizing for. Below is a structured comparison across the factors that matter most operationally.
| Factor | Fully Remote | Hybrid (2 to 3 days in office) | Full RTO (5 days in office) |
| Retention risk | Low. Attrition tends to track management quality more than location. | Moderate. Depends heavily on enforcement strictness and schedule design. | High. Senior ICs and high performers leave at elevated rates within one to two quarters, per the Chicago/Michigan and Baylor/Pitt research. |
| Hiring pool size | Largest. Access to global talent across all time zones. | Large, but limited to commuting distance or relocation-willing candidates. | Smallest. Restricted to a single metro area. |
| Time-to-hire | Fastest for distributed roles, especially with AI-assisted sourcing. | Moderate. Slightly slower due to location constraints on the candidate pool. | Slowest. Niche roles can take 90 or more days to fill locally. |
| Coordination overhead | Highest. Requires mature async processes, strong documentation, and intentional overlap windows. | Moderate. In-office days reduce some coordination cost but create calendar fragmentation. | Lowest for synchronous work, but eliminates time zone coverage advantages entirely. |
| Cost | Lowest. No per-engineer office footprint; many companies report significant compensation savings in global markets. | Moderate. Partial office footprint plus possible remote stipends. | Highest. Full real estate cost, utilities, and local-market salaries. |
Pros and Cons of Return to Office Mandates for Engineering Teams
Pros:
- Faster synchronous decision-making for ambiguous, high-stakes architectural choices
- Easier onboarding for junior engineers who benefit from proximity to senior mentors
- Reduced coordination overhead for teams that lack mature async processes
- Stronger informal culture-building through daily in-person interaction
- Simpler compliance and security posture for regulated industries or classified work
Cons:
- Elevated attrition among senior engineers and high performers within one to two quarters of a strict mandate
- Meaningfully shrinks the hiring pool, especially for location-bound postings
- Increases time-to-hire and backfill cost for departing engineers
- Adds 50 to 80 minutes of daily commute time, reducing work-life balance
- Eliminates follow-the-sun code review and on-call coverage advantages
- Increases real estate and facilities cost per engineer
- Correlates with measurable morale and satisfaction declines across independent studies, with no consistent evidence of offsetting financial or productivity gains
- Creates two-tier dynamics if some roles remain remote while engineering is mandated on-site
How Office Design and Schedule Design Shape Outcomes
If you choose hybrid, the physical workplace experience matters more than the policy document. Hot-desking without assigned neighborhoods frustrates engineers who need dual monitors and a consistent physical setup. Open-plan layouts without quiet zones undermine the deep-work benefit that is supposed to justify being on-site at all.
Companies reporting the best outcomes from hybrid engineering schedules tend to share three design choices:
- Designated team days. The entire squad or pod is in office on the same days, so in-person time is actually collaborative, not just co-located isolation.
- Protected focus blocks. At least one in-office day per week stays meeting-free, preserving deep-work time.
- Adequate workspace. Engineers have reserved desks with proper equipment, not shared hot desks that require setup and teardown every day.
In this context, culture is built by shared goals, mutual trust, and intentional rituals, whether those happen in a conference room or on a video call. The best hybrid policies treat office time as a tool with a specific purpose, not a default setting.
A Practical Decision Framework for Engineering Leaders
Before changing your work model, run through a structured evaluation. The cost of getting this wrong is measured in attrition, backfill expense, and months of lost velocity.
Questions to Ask Before Changing Your Work Model
- What problem are you solving? If the honest answer is “leadership wants people in seats,” that is not an engineering problem. Identify the specific output, collaboration, or quality metric you actually believe will improve.
- Do you have baseline data? Measure current deployment frequency, sprint velocity, attrition rate, and engagement scores before making a change. Without a baseline, you cannot tell afterward whether the mandate helped or hurt.
- Who will you lose? Survey senior ICs and staff engineers privately about whether a mandate would push them to look externally. If a meaningful share say yes, the likely backfill cost outweighs any collaboration gain.
- Can your hiring process absorb the attrition? If three to five senior engineers leave in the same quarter, does your recruiting team have the bandwidth and pipeline to fill those roles within 90 days?
- What does your async maturity look like? If your team already has strong documentation, clear PR review norms, and effective incident response across time zones, you are less likely to gain much from mandated co-location.
- Have you considered structured offsites instead? Quarterly in-person gatherings of three to five days often deliver most of the relationship and alignment benefit at a fraction of the cost and disruption of a full RTO mandate.
How to Structure a Hybrid Policy Without Creating Two Teams
The biggest risk of hybrid is an insider/outsider dynamic where in-office engineers get more visibility, faster promotions, and informal access to leadership. That dynamic drives attrition among remote team members and undermines the distributed hiring advantage you built in the first place.
To avoid it:
- Default to remote-first meeting practices. Even if half the team is in a conference room, everyone joins the call from their own laptop. This levels the communication playing field.
- Standardize async documentation. Every architectural decision, sprint outcome, and project update should be written down, not shared verbally to whoever happened to be in the office that day.
- Evaluate performance on output, not presence. Use engineering metrics (PRs merged, incidents resolved, features shipped) rather than desk visibility as the basis for feedback and promotion.
- Rotate on-site days if your team spans multiple offices. Alternate which location hosts planning sessions.
What to Measure After Rollout
If you do change your work model, track these metrics monthly for at least two quarters:
- Attrition rate by seniority level, watching for disproportionate senior departures
- Time-to-hire for open roles
- Deployment frequency and lead time for changes
- Employee engagement scores, via monthly pulse surveys rather than an annual one
- Meeting load per engineer per week
- Offer acceptance rate, watching for declines tied to location requirements
If attrition rises, time-to-hire increases, and deployment frequency stays flat or drops, the mandate is not working. Revisit within 90 days.
Frequently Asked Questions
Does return to office actually improve productivity?
Not in any way that shows up in the data engineering leaders track. Most teams that measure DORA metrics like deployment frequency and lead time haven’t seen consistent gains after an RTO mandate. Self-reported productivity is generally flat or slightly lower. What does reliably move is meeting volume on in-office days, and Microsoft and GitHub’s own DevEx research found deep, uninterrupted work, the thing offices tend to reduce, is the single strongest lever for individual output.
Do RTO mandates increase employee turnover?
Yes, and the effect is concentrated among the people companies can least afford to lose. A 2024 University of Chicago/Michigan study found strict mandates at Microsoft, SpaceX, and Apple pushed out senior employees at higher rates, mostly to competitors. Separate research from Baylor and Pittsburgh found a 13 to 14 percentage point rise in abnormal turnover after mandates, and Gartner found high performers are twice as likely as average employees to leave.
Which companies require 5 days in office in 2026?
Amazon and JPMorgan Chase are the two big names requiring five days for essentially their entire workforce. Most other major tech employers, Google, Meta, Apple, and Microsoft, have settled on three days instead. Shopify remains fully remote-first. Five-day mandates are still the exception, not the norm, even among large employers.
Is remote work still common for software engineers in 2026?
Yes. Stanford’s Survey of Business Uncertainty puts work-from-home at roughly 21.2% of all US paid workdays, and projects that figure will barely move even after every currently planned RTO mandate takes effect. Tech and engineering roles skew more remote and hybrid than the broader labor market.
Should engineering teams be hybrid or fully remote?
It depends on what you’re optimizing for. Fully remote maximizes hiring pool size and minimizes cost but requires mature async processes. Hybrid reduces some coordination overhead but risks calendar fragmentation and two-tier dynamics between in-office and remote staff. Full five-day RTO carries the highest retention and hiring risk with no consistent productivity payoff for engineering work specifically.
How strictly are RTO mandates actually enforced?
Less than the headlines suggest. Pew Research found 75% of workers were required to be in-office a set number of days as of late 2024, up sharply from 63% the year before, but enforcement varies widely. Some companies track badge swipes and issue warnings; many others state a policy without actively monitoring it, which is part of why actual office occupancy still lags well behind stated requirements.
Where This Leaves Engineering Leaders
The evidence points in one direction for distributed engineering teams: rigid RTO mandates carry measurable retention and hiring costs that most orgs underestimate, while the productivity gains they promise remain largely unproven for engineering-specific work.
If you’re building or scaling a distributed team and want to avoid the sourcing and vetting bottleneck that pushes some leaders toward “just get everyone in the office,” Arc’s global talent marketplace connects you with pre-vetted engineers across 190 countries, matched in days rather than months. See how Arc works for distributed engineering hiring.
