The 10 best Lenny's Podcast episodes
Episodes that will still be true in five years. Lenny releases weekly and a large share of the recent feed is about what AI did to product work this quarter — useful, and not what you go back to. These are the ten where a guest hands over a method.
These assume you already know what a PM does day to day and want the craft, not the introduction.
- 1The method you can adopt on Monday
A better way to plan, build, and ship products | Ryan Singer (creator of “Shape Up,” early employee at 37signals)
Ryan Singer explains Shape Up as an inversion: instead of estimating an idea, you fix an appetite — six weeks, sometimes two — and ask what meaningful version fits inside it. The shaping sessions, three to six hours with product, design and a senior engineer, exist to find the hidden branches before anyone is committed. It is the most directly implementable hour in the archive.
1h 45m Read the summary → - 2The one everyone quotes
Superhuman's secret to success: Ignoring most customer feedback, manually onboarding every new user, obsessing over every detail, and positioning around a single attribute: speed | Rahul Vohra (CEO)
Rahul Vohra on positioning Superhuman entirely on speed, ignoring most customer feedback on purpose, and onboarding every early user by hand. The buried detail is the org story: eight or nine direct reports left him 6-7% of his time for product, and hiring a president took that to 60-70%. The product-market-fit survey everyone borrowed came from this conversation.
1h 25m Read the summary → - 3The pricing hour
Pricing your AI product: Lessons from 400+ companies and 50 unicorns | Madhavan Ramanujam
Madhavan Ramanujam's argument is that pricing tests product-market-*pricing* fit, and you can run the test before a prototype exists. The mechanics are specific: index value against a reference product, probe acceptable/expensive/prohibitive thresholds to find the psychological cliffs, rank most and least valued features, talk to 20-30 accounts. Porsche validated every Cayenne feature this way before there were blueprints.
1h 12m Read the summary → - 4The shortest one worth an hour
4 questions Shreyas Doshi wishes he’d asked himself sooner | Former PM leader at Stripe, Twitter, Google
Forty-five minutes, four questions, no framework diagram. "Is this a must-win battle" and "what does great look like" are the two that do the work, and Doshi is specific about the cost: saying no to ninety percent of ideas to protect the other ten. Shorter than everything else on this list and denser than most of it.
46 min Read the summary → - 5The scaling one
The high-growth handbook: Molly Graham’s frameworks for leading through chaos, change, and scale
Molly Graham scaled Facebook from 500 people to 5,500 and Google's comms team from 25 to 125, and "give away your Legos" is the line that survived. The point underneath it is that a growing company keeps handing you more Legos than you can hold, so the job is continuously handing off the thing you have just got good at. Written from scars rather than from a framework.
1h 32m Read the summary → - 6The one about taste
Slack founder: Mental models for building products people love ft. Stewart Butterfield
Stewart Butterfield on utility curves — the S-shape where early effort buys nothing, the middle buys everything, and the tail buys complexity — as the tool for deciding whether to build the rest of a feature. Plus "divine discontent": standards rise with familiarity, so the core flows need reworking even when nobody complained. The Slack specifics (magic links, the shouty rooster on @everyone) are the evidence, not the point.
1h 31m Read the summary → - 7The one with 600 experiments behind it
Behind the product: Duolingo streaks | Jackson Shuttleworth (Group PM, Retention Team)
Duolingo's streak is the most copied retention mechanic in consumer software, and this is the team explaining what actually moved the number across 600-plus tests. Lowering the bar to a single exercise failed because it muddied what the streak meant; the wins came from clarity of copy and from days zero to seven, where loss aversion switches on. Rare in that the failed tests are in it.
1h 29m Read the summary → - 8The one that is not about product at all
How have I been complicit in creating the conditions I say I don’t want? | Jerry Colonna (CEO of Reboot, executive coach, former VC)
Jerry Colonna's question — how have I been complicit in creating the conditions I say I don't want — is the title, and he is careful about what complicit means: accomplice, not perpetrator. His example is the person who says they hate being busy and finds an empty calendar unbearable, because the busyness is quieting something else. The most uncomfortable hour in the feed and the one people come back to.
1h 23m Read the summary → - 9The founder one
Tobi Lütke’s leadership playbook: Playing infinite games, operating from first principles, and maximizing human potential (founder and CEO of Shopify)
Tobi Lütke on killing projects the moment he sees the flaw rather than letting them run — he calls it the Tobi tornado — and on why copying a working solution imports the constraints it was built under. The frank part is the arithmetic behind the bluntness: roughly forty working years per person, so leaving someone on a doomed project is spending a real fraction of it.
1h 42m Read the summary → - 10The career one
Building a long and meaningful career | Nikhyl Singhal (Meta, Google)
Nikhyl Singhal has mentored hundreds of PMs and the pattern he names is that people move jobs on a trigger — a bad manager, slow shipping — and land in the same situation with a new logo. His counter is to work backwards from the end state and treat this role as preparation for the job after next. The concrete warning is about "ex-growth" companies: high valuation, no natural pull, equity that evaporates.
1h 26m Read the summary →
What's missing, and why
The AI-era episodes. A large share of the 2025-2026 feed is smart people describing what AI is doing to product work right now — the Anthropic and OpenAI product leads, the state-of-AI conversations, the vibe-coding episodes. They are worth hearing the week they land and they are the first thing to age. This list is deliberately the other half of the archive; the recent ones are all on the show page.
Read one before you commit an evening to it
Every episode above links to its summary here — the argument, the chapter outline and the transcript, free and without an account. PodSized has 328 Lenny's Podcast episodes summarized; the full episodes stay in whatever podcast app you already use.
Frequently asked questions
What is the best Lenny's Podcast episode? ↓
How were these Lenny's Podcast episodes chosen? ↓
I have never listened to Lenny's Podcast. Should I start here? ↓
Why isn't "The AI-era episodes" on this list? ↓
Can I read these instead of listening? ↓
Keep going
Five entry points in order, and whether episode 1 is one of them.
More from Lenny's Podcast: Product | Career | Growth
- OpenAI’s Head of ChatGPT: We’re entering a new era of AI (again) | Tibo Sottiaux
- The grief, loneliness, and burnout sweeping through the tech industry right now | Molly Graham
- 90 minutes of unfiltered product advice from Snap and Discord’s product chief | Peter Sellis
- How we built Grok Bot in a month | Roman Ugarte (SpaceXAI)
- Why companies are becoming a series of loops | Anish Acharya (a16z)