Just attended 3 days of “Future of Work” panels at SF Tech Week and I’m processing some uncomfortable truths. The remote work debate isn’t over - it’s entering a new, messier phase. 
Events attended:
- CXAI “The Future of Work: Employee Experience, AI, and the Hybrid Revolution”
- WeWork “Who Will Decide the Future of Work?” panel
- a16z “Managing Distributed Teams” roundtable
- Y Combinator “Startup Culture in the Hybrid Era” session
My context: Engineering director, 45-person team, been doing hybrid (3 days in office) since 2023
The Data Bomb That Dropped
From CXAI panel - Fortune 100 research presented:
2023 vs 2025 office mandate comparison:
2023:
- Fully in-office: 5% of Fortune 100
- Hybrid: 78%
- Fully remote: 17%
2025:
- Fully in-office: 54% of Fortune 100 (!!)
- Hybrid: 41%
- Fully remote: 5%
That’s a 10x increase in full RTO in just 2 years.
Major tech companies (Oct 2025 status):
- Amazon: 5 days/week mandate (started Jan 2025, 350K employees)
- Google: 3 days/week (but Sundar said “stays flexible if productivity high”)
- Meta: 3 days/week for hybrid employees
- Microsoft: 3 days/week (starting Jan 2026)
- Apple: 3 days/week
- Salesforce: 3-4 days/week depending on role
The trend is clear: Remote is losing.
The Talent Exodus Reality
Data shared at WeWork panel (from Glassdoor analysis):
Companies with 5-day RTO mandates:
- Employee turnover: +13% higher than flexible companies
- Glassdoor rating drop: -0.5 stars average
- “Would recommend” drop: -12 percentage points
Employee response to 5-day RTO mandate (survey data):
- 44% would comply
- 41% would job hunt
- 14% would quit immediately (!!!)
8 out of 10 companies admitted losing talent due to RTO policies.
But here’s the twist (from a16z panel): Most of those employees who said they’d quit… didn’t. Because the job market in 2025 is tough.
Reality check: Employers have more leverage now than in 2021-2022.
My Team’s Experience (Real Numbers)
Our RTO journey:
2020-2022: Fully remote
- Productivity: High (shipped 3 major features)
- Morale: High (flexibility loved)
- Turnover: 8% (below industry average)
- Hiring: Easy (could hire anywhere)
2023: Hybrid 2 days/week
- Productivity: Stable
- Morale: Some complaints, mostly adapted
- Turnover: 12% (normal)
- Hiring: Still good (flexibility marketed)
2024: Hybrid 3 days/week (company mandate)
- Productivity: Slightly down (more time in meetings)
- Morale: Mixed (parents struggling most)
- Turnover: 18% (concerning)
- Hiring: Harder (candidates ask about remote first)
2025: Considering 4 days/week (CEO pressure)
- Team response: “Do it and we’ll start looking”
- Morale: Lowest I’ve seen
- Productivity: Actually UP (fear of layoffs)
- Hiring: Very difficult (losing candidates to remote-first companies)
The uncomfortable truth: My team is more productive out of fear, not because office mandates work.
The Productivity Paradox
The big question at every panel: “Does in-office actually improve productivity?”
The data is… inconclusive.
From CXAI research (analyzing Fortune 500 performance):
Companies with 5-day RTO vs. Companies with flexible policies:
- Stock performance: No significant difference
- Revenue growth: No significant difference
- Employee engagement: Flexible policies win by 22%
- Innovation metrics: No significant difference
Translation: Being in office doesn’t automatically make companies perform better.
But…
From a16z founder roundtable - startup data:
Early-stage startups (pre-Product Market Fit):
- In-person teams: 18% higher chance of reaching PMF
- Remote teams: 15% more likely to die (correlation, not causation)
- Hybrid teams: Middle ground
Growth-stage companies (post-PMF):
- No statistically significant difference in performance
- Remote actually slightly better (lower costs, global talent)
Pattern: In-person matters MORE when figuring things out, matters LESS when executing known playbook.
The Real Reasons for RTO (Unspoken)
What CEOs SAY (from panels):
- “Collaboration is better in person”
- “Culture needs physical presence”
- “Innovation happens in spontaneous hallway conversations”
What data actually shows (from WeWork panel economist):
Reason 1: Real estate commitments
- Many companies locked into 10-year leases
- Empty offices = wasted money = board questions
- Average: $500/month per employee for unused space
Reason 2: Management comfort
- 72% of executives prefer in-office (vs 32% of ICs)
- Managers feel more in control when they see people
- “Butts in seats” = work getting done (perception)
Reason 3: Performance management
- Easier to monitor in-office
- Harder to fire remote workers (documentation burden)
- 86% of CEOs said they’ll “reward” office attendance (KPMG survey)
Reason 4: Justified headcount reductions
- RTO mandate → voluntary attrition → no severance costs
- Amazon example: Some estimate 10-15% quit after 5-day mandate
- Quiet layoffs without calling them layoffs
Controversial take from one panelist: “RTO mandates are a backdoor layoff strategy.”
The Generational Divide
Data from We Work Remotely 2025 survey:
Gen Z (entered workforce during pandemic):
- 67% prefer remote or hybrid
- 41% would take 10% pay cut for full remote
- 58% say they’re more productive at home
- Major concern: “I don’t know how to network in person”
Millennials (established pre-pandemic):
- 72% prefer hybrid (not full remote)
- 35% miss office social aspects
- 52% say productivity same remote vs office
- Major concern: “I need boundaries between work and home”
Gen X / Boomers (leadership roles):
- 54% prefer in-office 4-5 days/week
- 62% believe in-person builds better culture
- 48% think remote workers are less committed
- Major concern: “I can’t tell if my team is working”
At the WeWork panel, a Gen Z founder said: “My biggest competitive advantage is offering remote work. I can hire senior talent from FAANG companies who don’t want to RTO.”
Boomer executive response: “You’ll learn why culture matters when you try to scale.”
The room got tense.
The Hybrid Trap
From a16z roundtable - “Why hybrid is hardest”:
Problems with hybrid:
1. The “proximity bias” effect
- People in office get more visibility
- Remote days = out of sight, out of mind
- Promotions favor office-goers (data: +15% more likely)
2. The meeting hell problem
- Hybrid means: Some in room, some on Zoom
- Zoom people are second-class citizens in discussion
- Result: Schedule all meetings for in-office days
- In-office days become 6 hours of back-to-back meetings
3. The coordination nightmare
- “Let’s sync on Tuesday” - half team not in office
- Core hours required → defeats flexibility purpose
- Social cliques form among people with same in-office days
4. The “worst of both worlds”
- Lost flexibility of remote
- Lost spontaneity of full in-office
- Added: Commute some days (still need to live close)
- Added: Maintain home office + desk in office
My experience: Hybrid is hardest to manage. It’s not a middle ground, it’s a different (harder) mode.
The Cost Analysis Nobody Talks About
From my own calculation (shared at roundtable):
Per employee annual cost:
Full remote:
- Salary: $150K
- Home office stipend: $2K
- Equipment: $3K amortized
- Total: $155K
Hybrid (3 days/week):
- Salary: $150K (same)
- Office space: $6K ($500/month allocated)
- Home office: $1K (still need some)
- Equipment: $3K (need both setups)
- Commute subsidy: $2K
- Total: $162K
Full in-office:
- Salary: $165K (need to pay more, less competitive otherwise)
- Office space: $10K ($833/month allocated)
- Equipment: $3K
- Commute subsidy: $3K
- Total: $181K
For a 50-person engineering team:
- Remote: $7.75M
- Hybrid: $8.1M (+$350K)
- In-office: $9.05M (+$1.3M vs remote)
ROI question: Is that $1.3M difference worth it for “better culture”?
Depends on who you ask.
What Actually Works (Real Examples)
From Y Combinator session - successful models:
Model 1: “Remote-first with offsites”
- Example: GitLab, Zapier, Automattic
- Daily: Fully remote, async-first
- Quarterly: Week-long team offsites
- Cost: ~$3K/employee/quarter for offsites
- Employee satisfaction: Very high
- Productivity: High (measured by output)
Model 2: “Hub-and-spoke”
- Example: Coinbase, Stripe
- Offices in 3-4 cities
- Teams choose their hub
- No mandate for specific days
- Cost: Lower than traditional (smaller offices)
- Flexibility: High (choose your hub)
Model 3: “Core days”
- Example: Atlassian, Salesforce
- Tuesday-Thursday required in office
- Monday/Friday flexible
- Everyone knows when team will be together
- Cost: Can reduce office space (not everyone Mon/Fri)
- Balance: Predictability + flexibility
Model 4: “Results-only” (most controversial)
- Example: Some YC startups
- Zero mandates
- Measured purely on deliverables
- Annual performance review decides if it’s working
- Cost: Lowest (minimal office)
- Risk: Requires strong performance culture
The YC partner’s take: “Remote works if your culture and processes work. If they don’t, RTO won’t fix that.”
My Biggest Takeaways
1. The “return to 2019” is impossible
People have restructured their lives:
- Moved to cheaper cities
- Took on caretaking responsibilities
- Built home offices
- Adjusted lifestyle
You can’t just undo 5 years.
2. One-size-fits-all policies fail
Different roles need different setups:
- Early-stage founders: More in-person helps
- Senior ICs: Remote is often better (deep work)
- Managers: Need some in-person (relationship building)
- Parents: Flexibility matters most
3. Trust is the real issue
If you don’t trust your team to work remote, you have a hiring problem or a management problem, not a location problem.
4. The market decides
In tight labor markets: Employees win (2021-2022)
In loose labor markets: Employers win (2024-2025)
Right now, employers have leverage. But that won’t last forever.
The Uncomfortable Prediction
From multiple panelists (consensus forming):
Next 2-3 years:
- More companies will mandate 4-5 days in-office
- Employees will comply (job market is tough)
- Talent will quietly leave when they can
- Companies with flexible policies will win the talent war
5-10 years:
- Pendulum swings back toward flexibility
- Younger generation enters leadership
- Companies that stayed flexible will have built remote-capable cultures
- Companies that forced RTO will struggle to compete for talent
One panelist: “RTO mandates in 2025 are optimizing for short-term control at the expense of long-term talent strategy.”
Another panelist (CEO): “We’re optimizing for survival. Long-term doesn’t matter if we don’t make it to long-term.”
Both are right.
What I’m Doing
My plan (shared this with my CEO):
Pushing back on 4-day mandate:
- Showed him the turnover data (18% is expensive)
- Showed him the hiring difficulty data
- Made the cost argument ($1.3M difference)
Alternative proposal:
- Keep 3 days/week hybrid
- Mandate Tuesday-Wednesday-Thursday (core days model)
- Make Monday/Friday no-meeting days for deep work
- Quarterly team offsites (add intentional culture time)
- Results-based performance (ship or don’t, location doesn’t matter)
His response: “Let me think about it. Board wants butts in seats.”
Translation: I’ll probably lose this fight.
My backup plan: Start building relationships with remote-first companies, just in case my team starts leaving.
Questions for the Forum
For other engineering leaders:
- What’s your company’s RTO policy?
- How is it actually going (honest assessment)?
- Have you seen productivity changes?
- What’s your turnover like?
For ICs:
- How many days/week are you in office?
- Would you leave for a remote role?
- Do you actually collaborate better in-person, or is it just meetings?
For founders:
- How did you decide your remote policy?
- Would you change it if you could?
Luis 
SF Tech Week - Future of Work panels
Sources:
- Fortune 100 RTO Analysis (July 2025)
- Glassdoor Employee Satisfaction Study (2025)
- KPMG CEO Survey (2025)
- We Work Remotely State of Remote Work Report (2025)
- a16z Startup Performance Data
