Lenny's PodcastFailure
CHAPTERS
- 0:00 – 0:30
Why this compilation is about failure—and what to listen for
Lenny frames the episode as a curated set of the most insightful failure stories from past interviews. He explains why failures often teach more than successes and invites feedback so he can do more compilation episodes like this.
- •Failure can be a faster teacher than success
- •This episode is a curated set of failure moments across past guests
- •Listeners are invited to share feedback and reactions
- 0:30 – 3:03
Sponsor break: Sendbird and Eppo (tooling to ship, test, and learn faster)
A short sponsorship segment highlights two products aimed at product and growth teams. The ads emphasize in-app communication (Sendbird) and experimentation velocity with deep analysis (Eppo).
- •Sendbird’s in-app communication platform (chat, bots, video, live)
- •Claim: in-app communication drives high conversion and satisfaction
- •Eppo as an A/B testing + feature management platform for modern growth teams
- •Emphasis on self-serve analysis and faster experimentation cycles
- 3:03 – 5:59
Katie Dill: The Airbnb design team “intervention” and the trust lesson
A newly hired design leader at Airbnb describes a jarring early failure: half the team and HR scheduled a meeting to read out what wasn’t working in her leadership. She shares how she resisted defensiveness, recognized the trust gap, and changed her approach to bring the team along.
- •Early leadership misstep: coming in “swinging” with changes too fast
- •The intervention-style meeting surfaced blunt, specific feedback
- •Core issue wasn’t each decision—it was lack of earned trust
- •Shift to listening, understanding motivations, and co-creating change
- 5:59 – 8:03
Rebuilding team engagement by leading with listening (and the measurable turnaround)
The story concludes with how the leader adjusted behaviors immediately—prioritizing listening and alignment while still pushing needed change. Engagement scores improved dramatically, reinforcing that change done with people beats change done to people.
- •Trust is a prerequisite for sustainable org change
- •Listening is a tactical leadership tool, not just a soft skill
- •“Inflicting change” vs. “building change together”
- •Improvement validated by engagement scores rising to best in company
- 8:03 – 11:15
Paul Adams: Freezing onstage at Cannes—and recovering in real time
Paul recounts a career-scarring public speaking moment while representing Facebook at Cannes: he froze mid-talk, walked offstage, and accidentally cursed while still mic’d. He explains how he regrouped, returned, and finished—proving survivability after worst-case failure.
- •Panic/freeze triggered by overly scripted, media-trained delivery
- •Walking offstage while mic’d created an amplified embarrassment
- •Regaining composure by accepting the moment and returning
- •Lesson: even public failures can be recoverable and temporary
- 11:15 – 14:22
Product failures at Google: building from fear (Buzz/Google+) and the cultural fallout
Paul shifts from stage failure to product failure, describing how Google’s social projects were motivated by competitive fear rather than user value. He contrasts this with Google’s successful products and explains how the fear-driven context shaped poor decisions and outcomes.
- •Buzz/Google+ were driven by competitive fear vs. user-centric purpose
- •Privacy issues and missteps compounded the failure
- •Strong companies can still have “bad neighborhoods” internally
- •Leadership lesson: fear-based strategy rarely yields great products
- 14:22 – 15:53
Leaving mid-project: being treated like a ‘spy’ and what it revealed about incentives
Paul describes leaving Google mid–Google+ effort for Facebook and the intense scrutiny that followed, including quarantine and laptop analysis. The episode uses this to underline how secrecy and fear can distort internal dynamics and trust.
- •Highly secretive project structure created internal antagonism
- •Departure was interpreted as betrayal because of competitive framing
- •Security scrutiny (quarantine, forensic review) signaled high paranoia
- •Root cause ties back to fear-based competition mindset
- 15:53 – 18:38
Embracing big bets: why failure reps matter and how to build ‘ship to learn’ culture
Paul argues that frequent failure is part of doing meaningful work, especially when making big bets. He shares Intercom’s cultural principles (“ship fast, ship early, ship often”) and the ongoing tension between high craft standards and rapid learning loops.
- •Failure is inevitable—and useful—when you take big swings
- •Culture can normalize iteration: ship early to learn and adapt
- •Hard trade-off: speed vs. quality/craft and fear of embarrassment
- •Advice: design the company to support experimentation
- 18:38 – 20:30
Tom Conrad: Why famous disasters can still accelerate your career (Pets.com + Quibi)
Tom reflects on working inside two iconic product/company failures and explains that these experiences can become “dominoes” that open doors later. He also notes how high-profile failures can be more legible to future employers than obscure successes.
- •High-profile failures can create unexpected career leverage
- •These experiences can build resilience and pattern recognition
- •Brand recognition (even negative) can make lessons transferable
- •Failure stories can be ‘timeless’ in what they reveal
- 20:30 – 23:21
Pets.com: Overfunding, arms races, timing, and the disciplined shutdown
Tom breaks down Pets.com’s dynamics: three heavily funded competitors spent irrationally on marketing, creating an unwinnable arms race. He challenges the simplistic ‘bad business’ narrative, emphasizing timing (dial-up era) and the decision to shut down and return remaining capital.
- •Excess capital can incentivize wasteful, competitive escalation
- •Timing mattered: the market/infrastructure wasn’t ready
- •The business model wasn’t inherently impossible (Chewy later proved it)
- •Leadership chose to wind down early and return funds rather than burn out
- 23:21 – 33:20
Quibi: Falling back in love with building—and then realizing the math was broken
Tom explains how Quibi rekindled his passion for product building, from forming teams to using modern architectures. But he concludes that the company’s underlying business equation couldn’t work within the investment/time horizon, regardless of execution quality.
- •Personal arc: returned to building after considering a career pivot
- •Huge production achievement: ~70 shows in 18 months
- •COVID disrupted daily-studio content plans and blurred differentiation vs. YouTube
- •Core lesson: if the business equation is broken, iteration can’t save it
- 33:20 – 35:52
Sri Batchu: “Fail conclusively” so failures become reusable learning
Sri reframes failure in growth work: the real failure is not learning. He argues experiments must be designed to produce conclusive outcomes; otherwise organizations repeat the same half-tested ideas whenever new leaders arrive.
- •Most growth experiments fail; culture must support risk-taking
- •Failure = lack of learning, not lack of revenue lift
- •Design tests to be decisive so the org doesn’t re-litigate ideas
- •In B2B, compensate for small sample sizes by maximizing treatment effect
- 35:52 – 39:01
How to test big go-to-market bets: maximizing treatment effect (ABM example)
Sri gives a concrete example with account-based marketing: teams often try ABM ‘halfway,’ get ambiguous results, and then repeat the same initiative later. His approach is to bundle tactics into a strong version first, then optimize down if it works.
- •Test the strategy (ABM) rather than small isolated tactics
- •Bundle multiple high-leverage tactics to create a strong test condition
- •If it fails under best effort, you can confidently stop
- •If it succeeds, iterate by removing costs and isolating what matters
- 39:01 – 42:10
JZ (Jiaona Zhang): Airbnb Plus—solution-first thinking, ops mismatch, and unit economics reality
JZ shares a major Airbnb product miss: Airbnb Plus, an attempt to solve trust/quality concerns by inspecting and ‘managing’ listings. She explains how competitor fear and solution attachment led to building an operationally heavy program that didn’t fit Airbnb’s strengths or unit economics.
- •Common PM trap: jumping to solutions and getting attached
- •Airbnb Plus tried to ‘manage inventory’ to raise trust/quality
- •Strategic mismatch: Airbnb lacked strong ops muscle for inspections
- •Unit economics and ‘magical thinking’ should be tested early
- 42:10 – 44:33
Alternatives to blunt solutions: represent homes better and solve targeted quality problems
JZ argues the underlying user need was knowing what you’re getting, which could be addressed through better representation and cheaper, targeted interventions. She highlights leveraging platform-native strengths (reviews) and choosing specific fixes (lockboxes, cleaning partnerships) over broad inspection.
- •Clarify the real problem: predictability and accurate expectations
- •Use scalable signals like guest reviews instead of paid inspectors
- •Target root causes per listing (e.g., access, cleanliness) with cheaper tools
- •Avoid one-size-fits-all programs when tailored solutions work better
- 44:33 – 52:23
Gina Gotthilf: The ‘A-side/B-side’ career—and why sharing B-sides builds resilience
Gina introduces the idea that people publicly share only their highlight reel, hiding the messy setbacks that actually shape careers. She shares personal lows—dropping out, visa issues, layoffs, firings—and emphasizes storytelling, persistence, and the long time horizon of a career.
- •Public narratives skew toward highlights; most careers are uneven
- •B-side moments include depression, dropping out, layoffs, visa loss, being fired
- •Lesson: perception and storytelling affect opportunity, not just achievement
- •Careers are long; setbacks at 26 can be minor in hindsight
- 52:23 – 57:53
Duolingo B-side product lessons: dogfooding, bad MVP experiments, and market misreads
Gina recounts internal mistakes at Duolingo: delaying a badges experiment, running an overly-lean test that couldn’t work, and forgetting to dogfood. She adds broader examples—failed social features, schools platform struggles, China blocking, and India localization assumptions—showing how easily teams can misinterpret data and context.
- •Bad experiment design can incorrectly ‘disprove’ good ideas (badges MVP test)
- •Dogfooding prevents obvious misses and should be institutionalized
- •Many initiatives fail; the key is to recover and integrate learnings
- •International expansion can fail due to regulation and incorrect local assumptions
- 57:53 – 1:00:28
Maggie Crowley: The best failure interview question—and a rewrite project horror story
Maggie shares her go-to interview question: ‘What’s the worst product you’ve ever shipped?’ because great PMs have failures and can own them. She tells a cautionary tale about a core product rewrite estimated at six months that dragged on for 2.5 years due to arrogance, skipped discovery, and poor upfront requirements work.
- •A strong PM can identify and admit their worst shipped product
- •Rewrites are high-risk and often massively underestimated
- •Skipping discovery and one-pagers leads to years of waste
- •Sunk cost fallacy keeps teams trapped long after warning signs
- 1:00:28 – 1:01:11
Wrap-up: gratitude for vulnerability and invitation to share your own failure story
Lenny closes by thanking guests for sharing vulnerable moments and reinforcing the theme that setbacks can be the catalyst for the next stage of growth. He invites listeners to comment or DM their own failure stories.
- •Failure can be the bridge to the next career/life step
- •Vulnerability makes the lessons usable for others
- •Call for listener stories and feedback
- •Episode ends with sign-off
