73% of Amazon Engineers Ready to Leave: Is Your RTO Policy a Talent Filter?

General
1

The numbers from the recent Amazon employee survey are staggering: 73% of engineers would consider leaving rather than comply with the 5-day RTO mandate. As an engineering director who’s navigated the transition from full-remote to hybrid, this doesn’t surprise me—but it should concern every engineering leader.

Here’s what keeps me up at night: we’re not just talking about losing average performers. The engineers most likely to leave are the ones with options—your senior ICs, your tech leads, the people who can land competing offers in a week.

The Flexibility Premium Is Real

Last quarter, I retained three senior engineers who had competing offers. All three offers included 15-20% salary increases. All three engineers stayed with us. Why? Our “core collaboration hours” model gives them what they actually value: autonomy over their time and location.

We require presence for specific moments—design reviews, quarterly planning, critical launches. But we trust them to optimize for productivity the rest of the time. One engineer told me: “I’d rather take less money and keep my Tuesday/Thursday focus days at home than commute 90 minutes for permission to wear headphones in an open office.”

The Diversity Impact Nobody’s Talking About

RTO policies disproportionately affect:

  • Women (who still handle 2x the household labor)
  • Parents and caregivers (who’ve built support systems around flexible schedules)
  • People with disabilities (for whom remote work removed significant barriers)
  • Geographically distributed talent (who relocated during pandemic with company approval)

When we mandate office presence, we’re not just filtering for “culture fit”—we’re filtering for privilege. Engineers who can afford to live near expensive offices. Engineers without caregiving responsibilities. Engineers whose disabilities don’t make commuting exhausting.

What Actually Happens When You Give Engineers Choice

Our team’s model: 40% fully remote (different time zones or states), 60% hybrid (2-3 days in office). Here’s what I’ve observed:

  • Productivity metrics (story points, PRs, incidents) show no correlation with office days
  • Best collaboration happens during intentional gatherings (quarterly offsites, design sprints)
  • Remote engineers are more likely to document decisions (which helps everyone)
  • Office days are social/relationship-building, not heads-down coding

The engineers who come in 4-5 days? They choose to. The engineers who come in 0-1 days? They’re equally productive. The difference is we’re not treating location as a proxy for commitment.

The Talent Market Has Shifted

Five years ago, flexibility was a perk. Today, it’s table stakes. The engineers I’m hiring—especially senior talent—have a new baseline expectation: “I’ll come in when it makes sense, not because it’s Tuesday.”

Companies forcing blanket RTO are competing with:

  • Fully remote companies (GitLab, Automattic, Zapier)
  • Hybrid-first companies (Atlassian, Dropbox, Airbnb)
  • Startups advertising “remote-first” as a recruiting advantage

When you mandate 5 days in office, you’re not competing for the full talent pool—you’re competing for the subset willing to accept that constraint. In a tight market for senior engineering talent, can you afford to filter out 73% of candidates before the interview?

My Question for Engineering Leaders

How are you thinking about flexibility as a competitive advantage in recruiting and retention?

Are you seeing similar data where engineers prioritize flexibility over compensation? How do you balance “culture” concerns with the reality that your best talent has options?

I’m genuinely curious how other leaders are navigating this. The old playbook—pay more, better benefits, interesting problems—still matters. But if you’re not offering flexibility, are you just selecting for engineers without better options?

2

Luis, thank you for naming the diversity impact explicitly. This is the conversation engineering leadership needs to have but keeps avoiding.

The Data We’re Ignoring

Our internal analysis showed that RTO policies don’t affect all demographics equally:

  • 68% of our women engineers cited caregiving as a flexibility need vs. 34% of men
  • Engineers with disabilities were 3x more likely to request permanent remote accommodations
  • Geographic diversity dropped 40% when we signaled “office-preferred” in job descriptions

The cruel irony? We spent years building DEI programs—inclusive interview practices, ERGs, sponsorship initiatives—only to implement policies that undo that work.

What “Culture” Really Means

When executives say “we need people in the office for culture,” I ask: which culture? The one that works for people who can afford to live near the office and don’t have caregiving responsibilities?

Because here’s what I’m seeing: our most inclusive, psychologically safe teams are the ones that built culture through intentional practices (clear documentation, rotating meeting times for time zones, async decision-making) rather than proximity.

The teams struggling with “remote culture”? They had culture problems before—they just used hallway conversations and after-work drinks to mask poor communication and unclear expectations.

The Retention Risk

Your point about losing engineers with options—that’s not random. It’s structural. The engineers most able to leave are often those least impacted by the barriers remote work removes:

  • They’re senior enough to negotiate
  • They don’t have caregiving constraints
  • They can afford to relocate for better opportunities

Meanwhile, the engineers who might stay despite preferring flexibility? They’re staying because they have fewer options—which means you’re selecting for people who can’t leave, not people who don’t want to.

That’s not a retention strategy. That’s a talent filter.

What Engineering Leadership Owes the Organization

We need to ask harder questions:

  • Is office presence a requirement for the work, or a preference based on how we used to work?
  • Are we measuring what actually drives performance, or using location as a lazy proxy?
  • What barriers are we creating, and for whom?

Flexibility isn’t just a recruiting advantage—it’s a DEI imperative. And if we’re serious about building diverse teams, we can’t implement policies that systematically exclude caregivers, people with disabilities, and geographically distributed talent.

3

The 73% number isn’t just about Amazon—it’s a leading indicator of how the talent market has structurally changed. And most engineering executives are making decisions based on 2019 assumptions.

The Leverage Shift

What’s different now: senior engineers have internalized that remote work is viable. They’ve spent 3+ years being productive from home. They’ve built lives around that flexibility. You can’t un-ring that bell.

This means your talent pool has segmented:

  1. Engineers willing to do full-time office (shrinking pool, self-selects for certain demographics)
  2. Engineers who want hybrid with real flexibility (largest pool, includes most senior talent)
  3. Engineers who will only consider remote-first (growing pool, includes specialized/senior ICs)

If you mandate 5-day RTO, you’re fishing in pool #1. If you offer “hybrid” but require 3+ days, you’re competing in pool #2 but losing to better hybrid models. If you’re not considering pool #3 at all, you’re missing 30-40% of senior IC talent.

What This Means for Strategy

I’ve been modeling this as a market positioning question: What talent segment do you need to win in, and what’s your differentiation?

For us, the calculation was clear:

  • We can’t out-pay FAANG on cash compensation
  • We can’t out-brand them on resume value
  • We CAN offer better flexibility, faster impact, more ownership

Flexibility became our moat. Not because we’re “nice,” but because it lets us compete for talent we’d otherwise lose to companies with deeper pockets.

The Real Cost of RTO

When I see companies announcing RTO mandates, I’m watching for:

  • Attrition of senior ICs (who have the most options)
  • Decline in geographic diversity of new hires
  • Increase in compensation required to close candidates (you’re compensating for the flexibility loss)

Amazon can weather losing 73% of engineers who’d consider leaving, because they have brand power and can hire replacements. Can your company? If you lose your top 20% of senior engineers, how long does it take to replace that knowledge and productivity?

The companies I’m watching closely—Gitlab, Zapier, Atlassian—aren’t doing remote/hybrid as a perk. They’re doing it as strategic differentiation in the talent market. And they’re winning talent from companies that treat flexibility as a “nice to have.”

4

Luis, your point about caregiving hit home. I’m one of those engineers who would have left if we’d mandated 5-day RTO.

The Reality of “Just Come to the Office”

My dad has early-onset Alzheimer’s. Three days a week, I handle his morning routine and get him settled with his caregiver. On remote days, I start work at 8am. On office days, I can’t get there before 10:30am—between the caregiving and the commute.

When people say “just come to the office for culture,” they’re not accounting for:

  • The 90-minute commute I’d need to add to caregiving time
  • The flexibility to handle medical emergencies without using PTO
  • The mental energy I’d lose to commuting that I currently put into my work

I’m a senior IC. I have options. And I’ve told my manager directly: I will leave before I give up the flexibility that makes my caregiving sustainable.

We’re Not Edge Cases

What surprised me: when I shared my situation in a team retro, four other engineers disclosed similar constraints:

  • One has a partner with a chronic illness
  • One is primary caregiver for aging parents
  • Two have kids with special needs requiring flexible schedules

All senior engineers. All would leave rather than lose flexibility. None had mentioned it before because they didn’t think the company would care.

Keisha’s point about DEI is critical: remote work didn’t create caregivers. It revealed how many of us have been quietly struggling to fit our lives into office schedules designed for people without caregiving responsibilities.

What Flexibility Actually Enables

I’m more productive now than I was pre-pandemic:

  • I code during my peak energy hours (early morning, late evening)
  • I don’t lose 2 hours a day to commuting
  • I can handle caregiving without the stress of “will I make it to the office on time?”

The irony: my manager initially worried I’d be “less committed” working remotely. Last quarter I shipped two major features, mentored three junior engineers, and had the highest PR velocity on the team.

Commitment isn’t location. It’s having the flexibility to do your best work while managing the rest of your life.

5

From a recruiting lens, the flexibility question has completely reshaped how we source and close engineering talent.

What Changed in Our Job Descriptions

Two years ago, our engineering JDs said “SF Bay Area preferred, remote considered.”

Last quarter, we changed to “Remote-first, office available in SF for those who want it.”

The result:

  • 3x more applicants per role
  • 40% increase in senior IC applicants (5+ years experience)
  • Geographic diversity up significantly (previously 80% Bay Area, now 45%)

We’re now competing for talent in Austin, Seattle, New York, and smaller tech hubs where cost of living is lower and engineers can afford houses.

The Closing Conversation

Here’s what I’m hearing in final interviews with senior candidates:

“What’s your remote policy?” used to be a question. Now it’s a filter. If the answer is “hybrid, 3+ days required,” 60% of senior candidates drop out.

The best candidates—the ones with multiple offers—are explicitly optimizing for:

  1. Compensation (still matters, but not the only factor)
  2. Flexibility (can they control their time and location?)
  3. Impact (can they ship meaningful work?)

If you can’t compete on #1 (because you’re not FAANG), you need to win on #2 and #3. Flexibility is how we compete for talent against companies with bigger comp bands.

The Talent Magnet Effect

Michelle’s point about market positioning is critical. We’re seeing talent concentration:

Companies offering real flexibility (not “you can WFH Fridays”) are attracting a disproportionate share of senior, specialized talent. Companies mandating office presence are competing for what’s left.

This creates a flywheel: better flexibility → attract better senior talent → ship better products → more attractive to next wave of talent.

The companies forcing RTO are betting they can hire replacements for the 73% who leave. Maybe Amazon can. Can you?

For us, flexibility isn’t charity—it’s competitive strategy. And it’s working.