We Stopped Hiring Junior Engineers—And Our Velocity Actually Increased

General
1

Six months ago, our EdTech startup faced a critical decision post-Series A: hire 15 senior engineers (L5+) or 30 junior engineers with the same budget. We chose the senior-heavy path—12 L5+ engineers and 3 solid L4s. Zero juniors.

I’ll be honest: the decision made me uncomfortable. As someone who started as a junior at Google, I’ve always believed in building diverse, inclusive teams across all experience levels. But the unit economics math was unavoidable.

The Results After Six Months

The data surprised even me:

Velocity metrics:

  • Deployed 2.3x more features with measurably higher quality
  • Technical debt actually decreased—seniors knew which shortcuts to avoid
  • Onboarding time dropped from 3 months to 3 weeks (seniors ramp faster)
  • Code review bottlenecks vanished (everyone could review anyone’s code)

The AI factor:
Our seniors paired with GitHub Copilot achieved roughly 3x the output of a junior engineer working alone. One L6 engineer with AI assistance could handle the workload we’d previously assigned to 2-3 junior developers.

VC perspective:
Our investors loved the story. Revenue per engineer increased 40%. Our burn multiple improved from 2.8x to 1.4x. We became a portfolio success case study.

The Uncomfortable Truth

But here’s what kept me up at night: our mentorship pipeline completely dried up.

Before this shift, we’d promoted 8 engineers from L2→L3 and 5 from L3→L4 over 18 months. We had a thriving culture of growth and development. Now? That’s gone.

When I talk to other VPs, I see the same pattern everywhere. We’re optimizing for this quarter’s velocity while potentially mortgaging our industry’s future. If everyone stops hiring juniors for 3-5 years, where will the next generation of senior engineers come from?

The Trade-off

Short-term: undeniable productivity gains, better unit economics, happier investors

Long-term: talent pipeline erosion, loss of diverse perspectives, potential industry-wide shortage in 5-7 years

My Question for This Community

Is this sustainable? Or are we collectively creating a crisis we’ll regret in 2030?

I want to believe we can find a middle path—maybe senior-led squads with 1-2 juniors each? Maybe formal apprenticeship programs? But right now, the economic pressure to hire senior-only is intense.

What are you seeing? What are you doing? And most importantly—what happens when we all wake up in 2031 and realize we haven’t trained anyone to replace today’s seniors?

2

Keisha, I’m nodding my head reading this. Your velocity gains are real, and I’ve seen the same pattern in financial services—but I want to push back gently on one thing: the pipeline risk you mention isn’t just a future problem. It’s already here.

The Talent Gap Is NOW, Not 2030

At my Fortune 500 company, we have 40 engineers. Our mix is 60% seniors (24 people), 25% mid-level (10), and 15% juniors (6). I fought hard to keep those 6 junior positions because I saw what you’re describing coming.

Here’s the uncomfortable data: average time-to-fill for senior technical roles is now 68 days. For AI/ML roles? 89 days. We’re already in a senior engineer shortage. And every company that goes senior-only intensifies that competition.

Juniors Fill Critical Gaps

AWS CEO Matt Garman made a point I keep thinking about: juniors are affordable, quick to adopt AI tools, and essential for long-term growth. In my experience, juniors handle work that seniors don’t want to touch:

  • Comprehensive testing and test automation
  • Documentation (which seniors “never have time for”)
  • Internal tooling and automation scripts
  • Support escalation and production debugging

When we tried shifting these to seniors? Morale dropped. Seniors wanted “interesting problems,” and maintenance tasks piled up.

The 2031 Question You Asked

“Who becomes your next senior engineers?” This keeps me up at night too.

If we don’t hire juniors for 3 years, that creates a missing cohort. In 5 years, you won’t have enough mid-level engineers. In 8-10 years, your senior bench is thin. The math is brutal.

As a first-gen college graduate from El Paso who started as a junior at Intel—if the 2026 hiring market existed in 2008, I wouldn’t be in this industry. That’s not theoretical. That’s thousands of talented people locked out.

My Hybrid Approach

I advocate for senior-led squads with 1-2 juniors each. Here’s the structure:

  • Each squad: 1 L5+ senior, 1-2 L3-L4 mid-level, 1 L1-L2 junior
  • Senior owns architecture, complex features, and mentorship
  • Mid-level owns feature delivery and code quality
  • Junior owns testing, automation, docs, and learning

Yes, it’s slower than all-senior. Probably 15-20% slower in raw feature velocity. But:

  • Knowledge transfer happens naturally
  • Bus factor decreases (what if that L6 leaves?)
  • Cultural health improves (growth mindset vs. hero culture)
  • In 18 months, that junior is a productive mid-level

The ROI Case

Let’s do the math on your scenario:

  • All-senior team: 12 engineers @ $170K avg = $2.04M/year
  • Hybrid team: 8 seniors @ $170K + 6 mid/juniors @ $110K = $2.02M/year

Similar budget, but now you’re building a pipeline. In 24 months, 2-3 of those juniors/mids are ready to step into senior roles. When your L6 engineers leave for FAANG (and they will—average tenure is 2.8 years), you have internal candidates ready to promote.

Question back to you and the community: In 5 years, when you need to backfill senior roles, would you rather promote from within or compete in an even tighter hiring market where senior salaries have inflated 40%?

I think the answer is clear, but it requires accepting short-term velocity trade-offs for long-term sustainability. That’s a hard sell to VCs focused on quarterly metrics, I know.

3

Both of you are describing a very real phenomenon, and I want to offer the strategic perspective from having navigated this exact decision at scale.

Team Composition Should Match Company Stage

Here’s the framework I use:

Pre-PMF (Seed, pre-Series A):

  • 100% senior engineers
  • Move fast, validate quickly, pivot without technical debt
  • No time for mentorship when survival is the goal

Post-PMF, pre-Series B:

  • Balanced mix: 60% senior, 30% mid-level, 10% junior
  • Building sustainable architecture while maintaining velocity
  • Start investing in culture and knowledge transfer

Series B+ (where we are now):

  • Pipeline investment becomes critical: 50% senior, 30% mid-level, 20% junior
  • Long-term talent development is competitive advantage
  • Knowledge distribution protects against key person risk

The Cultural Issue Nobody Talks About

Here’s what I learned when we tried the all-senior approach at our SaaS company in 2024:

We hired 18 L5+ engineers in 6 months. On paper, it looked amazing. In practice? Seniors competed for the interesting work.

  • Nobody wanted to own monitoring and observability
  • Documentation was “someone else’s job”
  • On-call rotations became political
  • Code reviews turned into architecture debates
  • Maintenance tasks created passive-aggressive slack threads

Juniors provide what I call “unglamorous reliability”—they’re excited to own monitoring, write docs, improve CI/CD. That work still needs to get done. When it’s all seniors, it either doesn’t happen or creates resentment.

The Cost Structure Warning

Luis’s math is spot-on, but there’s another risk: senior-heavy teams create unsustainable cost structures.

When the market shifts (and it will—we saw this in 2022-2023), senior salaries are the hardest to sustain. If you need to cut 30% of your engineering budget in a downturn, would you rather:

A) Lose 5 seniors, gutting institutional knowledge
B) Lose 8 juniors + 2 mid-level, keeping senior expertise intact

The all-senior strategy optimizes for best-case scenarios. But what’s your plan for the downturn that’s statistically likely to happen in the next 3-5 years?

My Recommendation

Hire for your stage + 18 months ahead, not for today.

If you’re Series A today planning Series B in 18 months, start building the Series B+ team composition now. That means accepting a 10-15% velocity trade-off today to avoid a talent crisis in 2028.

Keisha, you asked if this is sustainable. My answer: it’s sustainable for 12-18 months, maybe 24 if you’re lucky. But you’re trading short-term efficiency for long-term fragility. The question is whether you can afford to take the longer view—and whether your board will let you.

4

Coming from the design side, I’m fascinated by this conversation because we’re seeing the exact same pattern in design hiring. And I think y’all are missing something important about the value juniors bring.

Fresh Eyes See What Experience Misses

Here’s a story: Last year, my design team (all senior designers except one junior) was working on a complex B2B dashboard redesign. We spent 3 weeks iterating. Very sophisticated, very polished.

The junior designer—fresh out of bootcamp—asked what felt like a “dumb question”: “Why do users have to click 4 times to see their most important metric?”

We all paused. Because she was right. We’d optimized for visual elegance and information density (senior designer instincts) but lost sight of the primary user need. A junior, unburdened by our assumptions, saw the obvious problem.

Senior-only teams develop groupthink faster. Everyone has similar training, similar mental models, similar blind spots.

Diversity of Experience = Diversity of Perspective

When you’re building products—especially in EdTech, Keisha—you need people who:

  • Ask “why do we do it this way?”
  • Don’t know the “right way” yet, so they try unconventional approaches
  • Aren’t jaded by “we tried that before”
  • Bring current trends, new tools, fresh frameworks

Juniors provide that. Not always. Not consistently. But often enough to matter.

The EdTech-Specific Concern

Keisha, you’re in EdTech. You’re designing for learners. Don’t you want people on your team who remember what it’s like to be confused, to struggle, to not understand something obvious?

Seniors have 10+ years of expertise. That’s amazing for architecture and system design. But it also means 10+ years of forgetting what it’s like to be a beginner.

Your users are often beginners. Having junior engineers and designers who empathize with that experience is strategic, not charitable.

My Question

Are you optimizing for speed or for building the right thing?

Because if velocity was all that mattered, we’d all use the same tech stack, the same design patterns, the same solutions. But products win because they understand users better, not because they ship faster.

Maybe juniors slow you down. But if they help you build something 20% more valuable to users, isn’t that trade-off worth it?

5

This discussion is hitting on something that goes beyond engineering—it’s a product-market fit question disguised as a hiring question.

The Product Lens: Juniors Are a Customer Insight Engine

Maya’s point about fresh eyes is exactly right, and it has direct business implications. Let me share data from my fintech company:

Feature adoption rates:

  • Features designed by all-senior team: 35% adoption within 90 days
  • Features designed with junior input: 52% adoption within 90 days

Why? Juniors represent your early majority users—people who aren’t power users, who don’t have domain expertise, who experience your product with beginner’s eyes.

Your senior engineers think like senior engineers. They build for people like themselves. That’s fine if your customers are all senior engineers. But most aren’t.

The LTV:CAC Argument Nobody’s Making

Here’s the unit economics case FOR juniors that I haven’t seen mentioned yet:

Senior-only team:

  • Faster time-to-market: ships feature in 4 weeks
  • Feature works perfectly for 20% of users (power users)
  • Requires 3 iterations over 8 weeks to work for other 80%
  • Total time to market fit: 12 weeks
  • Customer acquisition suffers because early adopters churn

Balanced team:

  • Slower initial ship: 6 weeks
  • Junior asks “dumb questions” that surface usability issues early
  • Feature works for 70% of users on first launch
  • Requires 1 iteration over 3 weeks
  • Total time to market fit: 9 weeks
  • Higher early adoption, better retention, lower CAC

The math: Senior-only teams optimize for engineering velocity but miss product-market fit. That’s expensive.

The “Barbell Strategy” I Recommend

Don’t hire evenly across the spectrum. Go barbell:

70% seniors (L5+): Own architecture, complex features, technical strategy
30% juniors (L1-L2): Customer voice, usability testing, documentation, automation

Skip the middle. Mid-level engineers cost almost as much as seniors but don’t bring the fresh perspective of juniors or the deep expertise of seniors.

The Downturn Scenario Michelle Mentioned

She’s absolutely right. I lived through 2022-2023 layoffs. Companies with senior-heavy teams faced brutal choices:

  • Cut 25% of seniors → lost institutional knowledge, couldn’t recover
  • Cut 40% across all levels → kept capability but decimated morale

Companies with the barbell approach had flexibility:

  • Cut juniors first (hard but survivable)
  • Kept senior core intact
  • Rehired juniors when market recovered

Hiring isn’t just about filling today’s seats—it’s about building tomorrow’s team.

If you’re optimizing purely for Q2 2026 velocity, hire all seniors. If you’re building for 2030, you need a pipeline. The question is: what’s your board optimizing for?