LLenny's Podcast
← All episodes
Dalton Caldwell (Y Combinator, Managing Director)18 April 2024

Lessons from 1,000+ YC startups: Resilience, tar pit ideas, pivoting, more

10Frameworks
15Insights

Episode overview

Lenny Rachitsky interviews Y Combinator managing director Dalton Caldwell about the resilience behind successful startups and the practical meaning of his mantra, “just don’t die.” Caldwell explains how founders should assess whether to persist or stop, why strong pivots move toward accumulated expertise, and how deceptively attractive “tar pit” ideas trap teams. They also discuss customer conversations, early-stage delegation, investor decisions, startup failure, and Y Combinator’s Request for Startups.

Key ideas

  • Startup survival often depends on founders retaining hope, continuing high-quality attempts, and refusing to accept failure prematurely.
  • Persistence makes sense when founders still enjoy the work, their co-founders, customers, product, and problem; sustained misery and exhausted ideas can justify stopping.
  • Good pivots usually move closer to the founders’ expertise and build on knowledge acquired while pursuing the previous idea.
  • Tar pit ideas are especially dangerous because they sound compelling, attract repeated positive validation, and have nevertheless defeated startups for decades.
  • Founders should spend substantial time in direct customer conversations, overcome social discomfort, and help customers complete implementation after they say yes.
  • Early founders should remain close to product and customers and avoid over-delegating to senior hires whose big-company experience may not fit.
  • At the earliest investment stages, immediate market-size estimates can matter less than making something people want and demonstrating a credible path to users and growth.
  • Before writing code or fundraising, aspiring founders can validate an idea by talking to potential customers and attempting to pre-sell it.

Transcript available · source text is retained privately and is not published

Frameworks in this episode

People & resources mentioned

Attributed to the moment in the episode. Timestamps are approximate.

People · 20

  • Dalton CaldwellMentionsDalton is managing director and group partner at y combinator where he's worked for over 10 years across 21 different YC batches

    today my guest is Dalton Caldwell

  • Rujul ZapardeMentionsmentioned

    I knew rul really well he actually worked with me at at y see as a visiting partner

  • Lu ChengMentionsmentioned

    also Lou his co-founder credit to him too of course

  • Gustaf AlströmerMentionsGustav is a group partner at Y combinator where he's been for almost six years prior to that Gustav was at Airbnb

    one of your colleagues current colleagues not former colleagues Gustaf was on the podcast previously

  • Patrick CollisonMentions“there's this tweet by Patrick hollison that I've brought up a couple times on this podcast”

    Patrick was very Hands-On with all of his customers and was extremely available

  • Grant LaFontaineMentionsmentioned

    Grant from whatnot these are all very very different people

  • Tony XuMentionsthat was something that is ingrained in our culture from Tony Shu from our founder and CEO

    Tony from door Dash is so different than a lot of folks

  • Ryan PetersenMentionsat plexport his company even after it got huge he never became the generic corporate chief

    Ryan from flexport like these these are just very different personality types

  • Scott BelskyMentionsHe's a former founder starting a company called Behance that he sold to Adobe where he worked up the ranks to chief product officer

    this is exactly what Scott bsky shared in our episode

  • Mark ZuckerbergMentionsit was sort of a random decision for Mark Zuckerberg to move from a uh monthly active to a daily active

    folks like Zuck and Reed Hoffman Sam Alman Elon Sean Parker

  • Elon MuskMentionsWhether you like him or not, Elon, this is similar to his um if you watch him, he has sets these really ambitious goals

    folks like Zuck and Reed Hoffman Sam Alman Elon Sean Parker

  • Reid HoffmanMentionsSarah Tavel's former Greylock partner, credited with a maxim about strengths and weaknesses.

    I cold emailed Reed Hoffman when LinkedIn was like 12 employees and he just responded

  • Sam AltmanMentionsthere's everybody has uh and had and continues to have like such deep trust in in Sam and Greg and our leadership team

    when I met Sam he had dropped out a Stanford to work on Loop

  • Sean ParkerMentionshe was at the time I think the president

    I ended up giving Sean Parker a ride to the Oakland Airport

  • Owen Van NattaMentionsmentioned

    they recruited a new CEO um who was formerly the COO of Facebook called Owen vanada

  • Rupert MurdochMentionsmentioned

    rert murdo's got the juice he wants me to fix it

  • Marc AndreessenMentionsMark Andre had this visual on some podcasts have just imagined a 100,000 drones just coming out of China just at us.

    Mark andreon was on my board

  • Pete KazanjyCoinedPete is the author of my single favorite book on sales called Founding Sales.

    there's also this book called founding sales that I imagine you're familiar with uh Pete Kazi was on the podcast

  • Steve JobsMentionsSteve Jobs is just uh the bar he held for the company and for technical talent and for excellence was not uh wavering.

    it reminds me of a Steve Jobs quote

  • Lenny RachitskyMentionstoday we've got another very special compilation episode something I've been pulling on more and more with the podcast and the newsletter

    sure thing thanks so much Lenny I appreciate it

Resources · 50

  • Y CombinatorMentionscompany · Paul Graham and co-founders

    Dalton is managing director and group partner at y combinator

  • imeemCoinedcompany · Dalton Caldwell and co-founders

    Dalton was the co-founder and CEO of imim which was acquired by Myspace

  • App.netCoinedcompany · Dalton Caldwell and co-founders

    co-founder and CEO of app.net which was an early adsfree competitor to Twitter

  • Lenny's PodcastCoinedpodcast · Lenny Rachitsky

    welcome to the podcast

  • AirbnbMentionscompany · Brian Chesky, Joe Gebbia, and Nathan Blecharczyk

    let's talk about Airbnb obviously something you know a lot about

  • RetoolMentionssoftware · Retool

    cashew changed her idea and renamed to something called retool

  • BrexMentionscompany

    vond changed their idea and got really excited about it and renamed to brex

  • PostHogMentionssoftware · PostHog

    in the case of post hog pivoting into their idea they knew a lot about analytics

  • MixpanelMentionssoftware · Mixpanel

    they ended up PIV into something kind of like a mixed panel competitor

  • SegmentMentionssoftware · Twilio

    there was no Universe where they would have made up the idea for Segment

  • FlightCarMentionscompany

    he had done this Marketplace called flight car when he was younger

  • ZipMentionscompany · Rujul Zaparde and Lu Cheng

    they basically found out about all this procurement software and what the state-of-the-art was

  • Funko PopsMentionsproduct · Funko

    there was no one else trying to build marketplaces for Funko Pops

  • WhatnotMentionscompany

    there was no one else trying to build marketplaces for Funko Pops

  • KoalaMentionssoftware

    around the city on for square and koala back in the day

  • LocalmindCoinedsoftware · Lenny Rachitsky and co-founders

    I had this startup called local mind that uh allowed you to talk to people checked in in various locations

  • FoursquareMentionssoftware · Foursquare

    around the city on for square and koala back in the day

  • RazorpayMentionscompany

    I funded Razer pay which is I think the largest payment processor in India

  • GoogleMentionscompany · Google

    they did this amazing thing at Google and then you hire them

  • CodaUsessoftware · Coda

    I use Koda every day to help me plan each episode of this very podcast

  • Hacker NewsMentionswebsite · Y Combinator

    people on Hacker News were excited about it

  • LinkedInUseswebsite · LinkedIn

    they just called DMD people on LinkedIn and asked them for advice

  • StripeUsescompany · Patrick Collison and John Collison

    they would just sort of like install stripe into the customer's website

  • FlexportMentionscompany · Ryan Petersen

    Ryan from flexport like these these are just very different personality types

  • DoorDashMentionscompany · DoorDash

    Tony from door Dash is so different than a lot of folks

  • Request for StartupsMentionswebsite · Y Combinator

    recently you put out a request for startups which is essentially 20 categories of ideas that you want to fund

  • SpaceXMentionscompany · Elon Musk

    several of the folks that are actually going to space right now that aren't SpaceX are YC companies

  • LinkedInCoinedwebsite · LinkedIn

    I cold emailed Reed Hoffman when LinkedIn was like 12 employees

  • Homebrew Computer ClubMentionsevent

    it reminds me of stories about the homer Computer Club

  • LooptCoinedcompany · Sam Altman and Nick Sivo

    when I met Sam he had dropped out a Stanford to work on Loop

  • Boost MobileMentionscompany

    their customer was Boost Mobile

  • NapsterMentionssoftware · Shawn Fanning and Sean Parker

    this is actually before Shawn Parker went to Facebook he was part of Napster

  • PayPalMentionscompany · Max Levchin, Peter Thiel, and Luke Nosek

    Reed hoffen right he was a worked at PayPal right and then uh he did LinkedIn

  • MySpaceMentionswebsite

    basically I sold my first company in Myspace

  • Apple App StoreUsessoftware · Apple

    I had a lot of success in the App Store with imme

  • Andreessen HorowitzMentionscompany · Marc Andreessen and Ben Horowitz

    the major investor we had was Andre Horwitz

  • BurbnMentionssoftware · Kevin Systrom and Mike Krieger

    there was another portfolio company called bourbon which was originally a four square clone

  • HipstamaticMentionssoftware

    if you looked on the paid app store charts the number one app was hip stematic

  • InstagramMentionssoftware · Meta

    of course this is Instagram right and so it took off really quickly

  • PikPleaseCoinedsoftware · Dalton Caldwell and co-founders

    it's called pick please yeah

  • FacebookMentionscompany · Meta

    if you look at the Facebook story them getting 100% penetration on the Harvard campus first

  • Getting to YesRecommendsbook · Roger Fisher, William Ury, and Bruce Patton

    find the most popular sales books like getting to yes that everyone reads and just read those

  • Founding SalesRecommendsbook · Pete Kazanjy

    there's also this book called founding sales that I imagine you're familiar with

  • The WireRecommendstv show · David Simon

    I keep rewatching like The Sopranos and the wire and it always is different to me every time

  • The SopranosRecommendstv show · David Chase

    I keep rewatching like The Sopranos and the wire and it always is different to me every time

  • ColumboRecommendstv show · Richard Levinson and William Link

    I've been really enjoying watching old episodes of Columbo

  • SiPhox HealthUsesproduct

    I just signed up for and they do at home blood testing

  • Oura RingUsesproduct · Oura Health

    like my or ring and my Apple watch and all that good stuff like I've been a fan

  • Apple WatchUsesproduct · Apple

    like my or ring and my Apple watch and all that good stuff like I've been a fan

  • LennyPodcast.comMentionswebsite · Lenny Rachitsky

    you can find all past episodes or learn more about the show at Lenny podcast.com

Spot an error or want something removed? Request a correction or removal.

Insights & moments

The myth-busts, hot takes, explainers, and tools worth keeping.

Hot Take· 3

Hot Take29:00

Why TAM Barely Matters at the Earliest Stage

Dalton says market size matters more the later and higher-valuation the round, but at pre-seed it's near the bottom of his list. Uber, Airbnb, and Razorpay all had tiny TAMs at the start; he cares far more about whether founders can get users and are making something people want.

  • TAM is critical at later, high-valuation stages and less so the earlier you go
  • Uber, Airbnb, and Razorpay all had tiny TAMs at inception
  • Razorpay's TAM was tiny in 2015 because almost no one used credit cards in India, until the market 100x'd
  • At YC he worries about getting users, growth, and making something people want, not Excel-modeled market size

the earlier you go the less it matters and some of the most phenomenally good startups if you were really uh penic about it the…

Dalton Caldwell · 29:30
#market-size#tam#investing#yc
Hot Take48:30

The One Trait Behind Founders Who Build Huge Companies

Dalton says personality type doesn't predict success, he's seen quiet and extroverted founders both win. The common thread is an almost irrational internal conviction and self-belief so strong it bends the world, convincing employees and everyone around them that it will work.

  • No single personality type predicts success (Tony from DoorDash, Patrick Collison, Ryan from Flexport are all different)
  • The shared trait is that they deeply, internally want it and believe they'll make it work
  • Their conviction persists even when the data says it isn't working
  • That internal gravitational force warps the world and gets others to believe too

somewhere deep down in there they're like oh yeah I'm gonna make this work this is this company is gonna be big and they and…

Dalton Caldwell · 49:30
#founders#conviction#success-patterns
Hot Take1:05:30

Growth Hacking Is a Waste of Time for Very Early Startups

Dalton's contrarian take is that analytics, A/B testing, and growth-hacking are actively harmful for pre-product-market-fit startups. Most public startup advice is aimed at Series A/B companies, and seed founders who consume it get pulled away from the real work of landing and talking to a first customer.

  • Split testing and growth-hacking theory are for companies that already have scale
  • Much online advice is really aimed at Series A/B founders, not seed-stage ones
  • Consuming late-stage advice too early pulls founders away from getting a first customer
  • Study what your comparable companies did from zero-to-one, and ignore what they do today

doing all this analytics AB testing stuff um is a total waste of time for very early startups

Dalton Caldwell · 1:06:00

pay attention to Facebook when they were getting their first thousand users what were those tactics

Dalton Caldwell · 1:09:00
#growth-hacking#early-stage#contrarian#product-market-fit

Explainer· 3

Explainer14:30

A Good Pivot Is Like Going Home

Dalton explains that successful pivots get warmer, not colder, moving toward what the founders are already expert in and building on what they learned from the prior idea. Brex, Retool, PostHog, Zip, and Segment all pivoted into spaces they had accidentally gained deep expertise in.

  • A good pivot moves closer to something you have real experience in
  • It builds on insights gained while grinding on the earlier idea
  • Segment's founders could never have started with their final idea, they learned analytics by running the prior one
  • You don't need domain expertise before starting, it can come from trying to build the company

a good pivot is like going home you you know it's it's warmer it's closer to something that you

Dalton Caldwell · 15:30
#pivoting#product-market-fit#founders
Explainer23:30

What Makes an Idea a 'Tar Pit'

A tar pit idea isn't obviously bad, it's an idea many people independently arrive at, that seems like an unsolved problem, and that generates lots of positive validation, which is exactly what makes it dangerous. Classic examples are apps to coordinate hanging out with friends and Foursquare clones, ideas people have chased since the '90s.

  • A tar pit only qualifies because it seems like a good, unsolved problem
  • You get genuine positive feedback and even early users, which draws you in and gets you stuck
  • Examples: apps to coordinate meeting up with friends, music discovery, Foursquare clones
  • Dalton worked on a tar pit himself (music discovery); Lenny had one too (Localmind)

the the weird aspect of what we call a tarpit idea is an idea that a lot of people come up with and that it…

Dalton Caldwell · 24:30

part of being a true tarpit is that you can get good initial validation you get what I mean

Dalton Caldwell · 25:00
#tar-pit-ideas#ideation#startups
Explainer37:00

The Real Most Common Cause of Startup Death

Contrary to the fear of running out of money, Dalton says startups most often die because founders lose hope, resign themselves to failure, or have a falling-out. Usually there's still some money left, but the founders no longer have another move in them.

  • You can see it in a founder's eyes when they've resigned themselves to failing
  • Running out of firepower while feeling positive is rarer than founders think
  • More often founders still have some money but are out of ideas or fighting with co-founders
  • As long as you don't accept failure, there are usually more moves to try

it's way more common that they like their idea doesn't work and they have a big fight with their co-founder and then they can't agree…

Dalton Caldwell · 38:30
#startup-failure#resilience#co-founders

Story· 2

Story11:30

The Batch's 'Worst' Companies Became Brex and Retool

In the Winter 2017 batch, the two companies that looked objectively worst, a despondent VR headset team called Vyond and a struggling UK Venmo clone called Cashew, ended up becoming Brex and Retool. Dalton uses the story to show that mid-batch struggle tells you almost nothing about eventual outcome.

  • Vyond was a VR headset startup by Stanford dropouts who nearly shut down out of shame
  • Cashew was a UK peer-to-peer payments app that wasn't growing
  • Vyond pivoted and became Brex (a decacorn); Cashew pivoted and became Retool
  • The batch's seemingly worst companies became by far its most successful

vond changed their idea and got really excited about it and renamed to brex

Dalton Caldwell · 13:30

out of my 35 companies the ones that objectively seemed the worst in terms of like it's just everything is going bad we're by far…

Dalton Caldwell · 13:30
#pivoting#yc#brex#retool
Story45:30

The Collison Install: Closing Isn't Done Until It's Live

Stripe's founders would tell customers 'I'm in the neighborhood, I'll drop by,' then physically show up, take the keyboard, and install Stripe into the customer's site themselves so it actually went live. The lesson: a 'yes' means nothing if the customer never implements, so you must finish the last mile.

  • Customers often say yes then go quiet and never implement, guaranteeing churn
  • The Collison brothers used white-glove in-person installs to get Stripe live
  • They wouldn't leave until the implementation was actually shipped
  • Even after a yes, you're not done with sales until the last mile is complete

even when you get a yes you're not actually done with sales you have to finish the last mile to get the thing implemented

Dalton Caldwell · 47:00
#sales#stripe#onboarding#startups

Q&A· 3

Q&A09:00

How to Know When It's Actually Time to Give Up

Dalton gives a nuanced signal for when to quit: ask whether you're still having fun and still love your co-founders, product, and customers. If the startup is profoundly damaging your life and relationships and you no longer care, that's a reason to stop. He reassures founders that shutting down carries little long-term shame.

  • Still having fun and enjoying your co-founders is a signal to keep going
  • Founders who turn it around genuinely love their customers, product, and each other
  • If it's profoundly hurting your relationships and mental health, leaning toward stopping is reasonable
  • In 10-20 years no one remembers that you shut a company down if you handled yourself with integrity

one are you still having fun do you still enjoy doing what you're doing

Dalton Caldwell · 09:00

no one will remember that you that you shut down your company probably in 10 years or 20 years

Dalton Caldwell · 11:00
#quitting#mental-health#founders#decision-making
Q&A17:30

When to Pivot: Count How Many Growth Ideas You Have Left

Dalton's heuristic for deciding whether to pivot is to look at how many untried ideas the founder still has for making the business grow. Running out of ideas is a signal to pivot; still having half a dozen good untested growth ideas is a signal to stay the course, as Airbnb did with its many zany experiments.

  • Out of ideas and not growing = usually a good time to pivot
  • Still having many untried, credible growth ideas = a reason to stay the course
  • Airbnb tried cereal, conventions, and many zany growth ideas without running out
  • Weak ideas like 'maybe we should pay influencers' can signal it's time to pivot

if it's not going well and you're out of ideas that is usually a good time to Pivot

Dalton Caldwell · 18:00
#pivoting#growth#decision-making
Q&A26:30

Why Investors Really Say No

Dalton argues founders should put themselves in an investor's shoes: investors make very few bets and are waiting for the ones they're most excited about. A 'no' usually just means it didn't clear that high bar, not that there's a hidden flaw in your pitch, Zoom setup, or shirt color.

  • Investors have limited budget and only make a few investments
  • They're waiting for the ones they're personally most excited about or think can be enormous
  • Anything that isn't clearly 'the one I want to do' gets a no
  • There's rarely a secret hidden reason, the feedback really is just 'we didn't want to invest'

anything that doesn't seem like this is the one this is the one I want to do is a no

Dalton Caldwell · 28:00
#fundraising#investors#founders

Takeaway· 4

Takeaway06:30

The One Mantra: Just Don't Die

Dalton's most repeated and most impactful advice to founders is simply to keep going and not let the startup die. Looking across every YC company, the underlying theme is that rationally the founder should have given up at some point but irrationally kept going through repeated near-death experiences before looking like an overnight success.

  • The most impactful advice isn't a clever move, it's the constant reminder to keep going and do high-quality reps
  • Rationally, most successful founders 'should' have given up at some point
  • Airbnb probably should have shut down three or four times before getting into YC
  • Success usually comes after multiple near-death experiences, then luck, then the 'overnight success' narrative

one of my mantras is just just don't die just keep your startup going just keep going

Dalton Caldwell · 06:30

the underlying theme is that rationally the founder should have given up at some point

Dalton Caldwell · 07:00
#resilience#startups#founders#yc
Takeaway19:30

Find a Big Incumbent Customers Hate (The Zip Playbook)

When Zip's founders had great execution instincts but no clear market, Dalton suggested finding large, publicly-traded or PE-owned companies that customers hate, then attacking that market with better software. Zip applied this to procurement software and it worked exceptionally well.

  • Look for large incumbents that are publicly traded or PE-owned and hated by their customers
  • Pair a knowable big market with an incumbent whose software is terrible
  • Zip used this exact prompt to land on procurement software
  • Effectively: find a large incumbent with very low NPS and try to disrupt them

start by looking at what companies are um publicly traded uh and or owned by private Equity that are large and that also are hated…

Dalton Caldwell · 20:00
#ideation#b2b#zip#market-selection
Takeaway32:00

You Can't Delegate Caring About Your Users

A recurring basic mistake is over-delegating and hiring impressive, senior executives with shiny resumés too early, often under investor pressure to scale. The best product founders stay deeply in the weeds and keep talking to customers no matter how late-stage the company gets.

  • Don't over-delegate; founders must stay close to product and customers
  • Watch out for hiring fancy-resumé senior people too early
  • Investor pressure to spend money and 'look serious' drives premature executive hires
  • If you genuinely care about customers and product, your instincts on where to spend time are good

you can't delegate caring about your users and you can't delegate caring that the product is great

Dalton Caldwell · 33:00
#hiring#delegation#product#founders
Takeaway40:30

Talk to Customers in Person, and Audit Your Calendar

Everyone 'knows' to talk to customers but few actually do it, often because of social anxiety about approaching people in the real world. Dalton offers a concrete self-assessment: count your in-person customer meetings this month, and aim for roughly 20-30% of your calendar to be customer conversations.

  • Hiding behind a keyboard, landing pages, and Instagram ads isn't talking to customers
  • The real blocker is usually social anxiety and fear of looking stupid
  • Self-assessment: how many in-person meetings with potential customers did I have this month?
  • Heuristic: about 20-30% of your calendar should be customer meetings and calls

you can do a self assessment in the past month how many in-person physical meetings have I had with potential customers

Dalton Caldwell · 42:00

there should be you know 20 or 30% of your time that the calendar says something like customer meeting customer call like

Dalton Caldwell · 43:00
#customer-discovery#sales#founders