Lenny's PodcastThe engineering mindset | Will Larson (Carta, Stripe, Uber, Calm, Digg)
CHAPTERS
- 0:00 – 0:34
Why leaders should stop “coddling” engineers
Will opens with a core belief: organizations often treat engineers like children, shielding them from real problems. He argues that accountability and exposure to hard trade-offs are what enable engineers to grow into true senior leadership.
- •Treat engineers as peers, not as fragile specialists
- •Sheltering engineers from “important work” limits growth
- •Accountability is a prerequisite for senior technical leadership
- •The shift in the market creates an opportunity to reset expectations
- 0:34 – 4:03
Meet Will Larson: CTO, author, and prolific engineering writer
Lenny introduces Will’s background across Carta, Stripe, Uber, Calm, and Digg, plus his books and blog. He previews the conversation topics: strategy, PM/EM relationships, writing, productivity metrics, values, and a Digg failure story.
- •Will’s leadership roles across major tech companies
- •Books: An Elegant Puzzle, Staff Engineer, upcoming Engineering Executive’s Primer
- •Will’s blog as a long-running source of engineering leadership insights
- •Episode roadmap: strategy, writing, productivity, values, and failure corner
- 4:03 – 6:27
Engineering in the post-ZIRP era: hiring slows, restructuring rises
Will contrasts the steady growth of the last decade with today’s volatile market. Leaders who were rewarded for hiring at scale now need to demonstrate different skills—team sizing, consolidation, and deeper operational leadership.
- •From frequent hiring to minimal interviewing and hiring freezes
- •Past success metrics (hire/retain) don’t map cleanly to current reality
- •Managers must shift from growth-mode to allocation and efficiency
- •Team consolidation and cuts become a core leadership competency
- 6:27 – 8:33
Leverage shifts—and the upside: holding engineers accountable enables growth
Lenny and Will discuss how reduced engineer leverage can remove incentives to over-coddle. Will frames accountability as positive: it enables trust, delegation, and placing engineers into genuinely senior roles.
- •Retention-driven management can lead to unhealthy “coddling”
- •Accountability unlocks senior responsibility for engineers
- •Staff+ roles require expectations, ownership, and consequence
- •Market dynamics may be pushing orgs toward healthier balance
- 8:33 – 13:14
Systems thinking: powerful tool, dangerous when it replaces reality
Will explains systems thinking through stocks and flows, emphasizing that mental models are always imperfect. He shares a Stripe incident-management example where the team over-optimized measurement instead of making improvements.
- •Systems thinking = modeling stocks (accumulations) and flows (rates of change)
- •Models are learning tools; reality is the ultimate arbiter
- •Common failure mode: prioritizing analysis/measurement over action
- •Recommended reading: Donella Meadows’ Thinking in Systems
- 13:14 – 16:23
Applying systems thinking to hiring: diagnose bottlenecks with a pipeline model
Will translates stocks-and-flows into a hiring funnel model: sources, stages, conversion rates, and constraints. He shows how simple pipeline visibility highlights whether the issue is sourcing, decision-making, or closing offers.
- •Start with any model; refine it against real data (ATS history)
- •Identify where candidates accumulate and where conversion collapses
- •Typical bottlenecks: hiring manager conviction vs offer acceptance
- •Avoid large process changes driven by vibes instead of evidence
- 16:23 – 18:52
Engineering strategy: you already have one—writing it down makes it debuggable
Will argues that most organizations do have strategies, but they’re often implicit and inconsistent. Writing strategy down enables debugging: is execution wrong, is interpretation wrong, or is the strategy itself wrong for a given context?
- •“No strategy” usually means “no written strategy”
- •Writing exposes inconsistencies and misunderstandings
- •Strategy documents enable diagnosis and iteration
- •Good strategy is often boring but clarifying
- 18:52 – 21:42
Boring constraints as strategy: standard kits, fewer tools, sharper focus
Will describes effective engineering strategies as constraints that force focus on what matters. Examples include restricting tool sprawl and standardizing on an existing “kit,” even if it frustrates some engineers.
- •Constraints reduce decision noise and concentrate effort
- •Standardizing tools trades individual autonomy for org leverage
- •Painful alignment can be worth long-term speed and cohesion
- •Strategy should optimize for company priorities, not universal happiness
- 21:42 – 25:08
Concrete strategy examples: Uber’s no-cloud rule and Stripe’s Ruby monolith
Will shares how Uber’s data-center-only approach enabled rapid expansion under geopolitical constraints, and how Stripe’s Ruby monolith focused engineers on product value rather than platform diversity. Lenny highlights the common thread: purposeful constraint.
- •Uber: strict no-cloud policy (at the time) and rapid China build-out
- •Stripe: Ruby monolith reduced tooling fragmentation and boosted focus
- •Good strategy allocates limited capacity to the highest-value problems
- •Bad strategy often stems from an inaccurate (or willfully ignored) diagnosis
- 25:08 – 26:48
How to get better at strategy: improve diagnosis with models and reading
Will recommends learning strategy fundamentals from Richard Rumelt and pairing that with systems thinking to sharpen diagnosis. He also calls out the lack of a single canonical “engineering strategy” text and hints at writing more in this space.
- •Start with Rumelt’s Good Strategy/Bad Strategy (then The Crux)
- •Use systems thinking to model reality and diagnose constraints
- •Additional reads: Technology Strategy Patterns, Value Flywheel Effect, Phoenix Project/The Goal
- •Engineering strategy literature is still fragmented—room for better canon
- 26:48 – 35:23
Writing what energizes you: why schedules and trends burn creators out
Lenny and Will align on a key principle: sustained creation requires intrinsic interest. Will explains why he avoids writing for deadlines or money and why quitting too soon is the biggest risk in content creation.
- •Write what you genuinely want to explore—energy is the fuel
- •Trend-chasing converges creators onto crowded topics (crypto → AI)
- •The biggest risk is burnout and quitting, not slow initial growth
- •Create for the long game; audiences compound over years
- 35:23 – 37:41
Making time to write with a demanding job (and a kid): align writing with work
Will shares practical tactics for maintaining output despite limited time. The core move is aligning writing topics with real work problems so writing strengthens job performance instead of competing with it.
- •Pick topics that overlap with current responsibilities and decisions
- •Life constraints (kids) force schedule redesign: nights vs weekend blocks
- •Energy management is the real bottleneck, not just calendar time
- •If you’re excited, you’ll find time; if not, the plan collapses at 9pm
- 37:41 – 41:18
Advice for aspiring writers: build artifacts or publish consistently—choose one
Will distinguishes between writing for career leverage and writing as a long-term practice. For the former, craft a few high-quality artifacts; for the latter, publish often and don’t over-optimize polish or audience reaction.
- •Career-focused: write 2–3 excellent pieces with heavy revision and feedback
- •Consistency-focused: publish frequently; don’t hoard drafts
- •Don’t “debug people” on the internet—ignore low-signal criticism
- •Books can be the polished synthesis; blogs can capture learning in motion
- 41:18 – 48:24
PM–Engineering partnerships: incentives, empathy, and shared performance ratings
Will explains two common EM/PM failures: misaligned incentives and conflict without understanding. He advocates for deeply surfacing true needs—and shares a structural fix: pairing PM/EM performance outcomes to force shared accountability.
- •Two root causes: incentive misalignment and misunderstanding the other role’s needs
- •Start by understanding before trying to solve the conflict
- •Invisible constraints: EMs often manage “interesting work” pressure from engineers
- •Organizational fix: calibrate PM/EM together; often give the same performance rating
- 48:24 – 55:36
Measuring engineering productivity: use metrics to diagnose and educate, not judge
Will addresses pressure to “prove” efficiency, especially to boards and investors. He recommends combining narrative proof of impact with diagnostic metrics (e.g., DORA), and using imperfect measurements as a vehicle to educate stakeholders.
- •Benchmarks can placate boards but don’t run the org effectively
- •Talk to engineers—teams often know where effectiveness breaks down
- •Show a credible list of meaningful outcomes delivered over the last 6 months
- •DORA/Accelerate metrics are great for diagnosis; dangerous as simplistic judgment tools
- 55:36 – 1:02:14
Defining values that work: honest, applicable, and reversible
Will critiques “cargo-cult” values copied from other companies and identity statements that can’t guide decisions. He offers a practical rubric: values must reflect reality, be usable in day-to-day trade-offs, and be meaningfully reversible.
- •Avoid copying values (e.g., writing Facebook’s values on your walls)
- •Honesty: values must describe what you actually do
- •Applicability: values should guide real decisions and trade-offs
- •Reversibility: if the opposite is unthinkable, it’s likely a useless identity value
- 1:02:14 – 1:11:15
Failure corner: the Digg V4 rewrite—death march, outages, and hard-won growth
Will recounts the Digg V4 rewrite: leadership turmoil, a high-stakes launch, prolonged instability, and heroic debugging under pressure. Despite eventual technical recovery, the business continued to collapse—yet the experience shaped his career trajectory.
- •Rewrite to add social features amid competitive and SEO-driven decline
- •Launch led to major instability; partial recovery took weeks
- •Key debugging story: Python default-parameter gotcha created escalating load
- •Career impact: accelerated responsibility and learning in a “rough time” company
- 1:11:15 – 1:16:53
Upcoming book and lightning round: recommendations, interviewing, and mottos
Will plugs The Engineering Executive’s Primer and explains who it’s for—current and aspiring execs and cross-functional partners. In rapid-fire answers, he shares favorite books, a candidate-closing interview question, and decision-making heuristics.
- •Book: The Engineering Executive’s Primer (O’Reilly) and where to find previews
- •Lightning books: Thinking in Systems, Good Strategy/Bad Strategy, Don’t Think of an Elephant
- •Interview question: how candidates choose among multiple offers (reveals values)
- •Decision heuristic: “Will anyone remember this decision in six months?”
