TThe Tim Ferriss Show
← All episodes
25 January 2023

#651: Legendary Investor Bill Gurley on Investing Rules, Finding Outliers, Insights from Jeff Bezos and Howard Marks, Must-Read Books, Creating True Competitive Advantages, Open-Source Strategies, Adapting Mental Models to New Realities, and More

10Frameworks
14Insights

Frameworks in this episode

Insights & moments

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

Explainer· 4

Explainer01:30

Why Gurley calls Porter's Competitive Strategy the most efficient short-form MBA

Gurley opens by defending Michael Porter's Competitive Strategy against Silicon Valley's reflexive contempt for MBAs. He estimates 80 or 90 percent of the entrepreneurs he meets would benefit greatly from reading just the first three chapters. His diagnosis of the most common founder mistake is precise: they arrive with a technological breakthrough in their own mind but spend no time analysing industry structure or whether the go-to-market is even possible. Porter supplies a framework for the competitive dynamics of an industry — whether you can break in, succeed, and then maintain that success. The book recurs throughout the episode as the substrate under Gurley's later points on switching costs, substitutes, moats and regulatory capture.

  • Competitive Strategy is the first book assigned in a first-semester corporate strategy course.
  • Silicon Valley dismisses the MBA, yet most founders would gain from the first three chapters.
  • The common failure is analysing your own technology and never the industry structure.
  • Porter frames whether you can break into an industry and hold position once in.
  • Gurley pairs it with Innovator's Dilemma and Crossing the Chasm as the startup canon.

one of the most common mistakes entrepreneurs make is they come up with some kind of technological breakthrough in their own mind but they don't…

Bill Gurley · 02:30

I would say 80 or 90 of the entrepreneurs I meet would benefit greatly by reading the first three chapters of this book

Bill Gurley · 02:00
#competitive-strategy#industry-structure#founder-mistakes#business-books#moats
Explainer56:00

Android, Kubernetes and Open Compute: open source used defensively

Gurley describes a shift over the last 10 to 15 years in how open source gets used. Android was Google's answer to an iPhone that scared the handset makers and telcos as much as it scared Google: build a competitor, make it open source, and the world lines up behind you. Facebook's Open Compute Project defines open standards for data centre hardware, commoditising its suppliers and neutralising IP claims. Google gifted Kubernetes to a Linux Foundation consortium and recruited IBM and HP, so nobody would be locked into AWS — eventually forcing Amazon itself to support it. He also notes open source suits complex problems better than simple ones, that Linux is arguably the most secure OS precisely because it gets beaten up the most, and that the Linux Foundation acts as a neutral steward doing patent defence.

  • Around 10 to 15 years ago companies began using open source defensively, not offensively.
  • Android was a defensive answer to the iPhone that also reassured telcos and handset makers.
  • Open Compute commoditises Facebook's hardware suppliers and disarms IP claims.
  • Kubernetes was gifted to a consortium so nobody would be locked into AWS.
  • Open source handles complex problems better than simple ones — Linux being the case in point.
  • The Linux Foundation acts as neutral steward, breaks ties and runs patent defence.
  • Gurley argues autonomous vehicles and NIH-funded research should be open source.

the thing is like open source is way better complex problems than simple problems

Bill Gurley · 57:30

people started using open source in more complex ways and sometimes using it defensively rather than offensively

Bill Gurley · 58:30
#open-source#defensive-strategy#kubernetes#platform-competition#standards
Explainer1:11:30

The unlearning thread: why the 2022 reset broke a generation's mental models

Gurley walks Ferriss through his spring 2022 tweet thread on valuation. The setup is that an entire generation built its views during the back half of a 13-year bull run, and the unlearning would be painful. He describes the cycle as a sawtooth rather than a sine wave: risk-on accumulates slowly like a boiled frog, then the crash happens all at once, roughly every seven to fifteen years. Because the run from 2009 lasted so long, many founders had never seen a reset, and the era taught them not just wrong valuation numbers but growth at all costs and the assumption you can always raise in nine months. His four points: forget previous all-time highs, treat revenue multiples as a crude proxy, expect to be valued on free cash flow and earnings, and understand revenue and earnings quality.

  • A generation formed its valuation intuitions during the back half of a 13-year bull market.
  • Risk-on builds slowly; the crash happens all at once, so the cycle is a sawtooth.
  • Down 70 percent does not mean cheap — previous highs are irrelevant.
  • Revenue multiples are a hack proxy; 10x should be an amazing upper limit, not a floor.
  • Founders anchored their net worth on peak valuations, which is psychologically destructive.
  • Startup failure rates were likely at all-time lows because money was so easy.
  • The fastest recovery is adjusting your mental models rather than waiting for normal.

previous all-time Highs are completely irrelevant it's not quote unquote cheap because it is down 70 percent forget those prices happen

Tim Ferriss · 1:12:00

risk on is a lot like the boiled frog like you don't know what's happening

Bill Gurley · 1:14:00

this is normal dude like that was a fantasy you were in and you need to forget it fast

Bill Gurley · 1:16:00
#valuation#market-cycles#unlearning#mental-models#venture-capital
Explainer1:48:30

The three books Gurley gifts most, and the case for far analogies

Gurley names his most-gifted books. Complexity by Mitchell Waldrop, on the rise of the Santa Fe Institute, is the one he has given away most and the book that introduced him to network effects. Mr China is a fund manager's humorous account of getting his head handed to him privatising Chinese industries in the mid-90s — his version of Red Notice, without the death threats. The third is David Epstein's Range, a counterpunch to the 10,000-hour rule, which he has been fascinated by for four or five years. The idea he extracts is that many of the biggest scientific breakthroughs come from people who changed disciplines and arrived with a different mental framework. He connects it to UCLA's Holyoak on far analogies — treating the ability to borrow ideas from far away as an intellectual skill — and to Santa Fe's multidisciplinary mix.

  • Complexity by Mitchell Waldrop is his most-gifted book and his route into network effects.
  • Mr China documents a foreign investor getting wrecked in 1990s China.
  • Range counters the 10,000-hour thesis in favour of breadth.
  • Big scientific breakthroughs disproportionately come from discipline switchers.
  • Holyoak at UCLA treats far analogies as a trainable intellectual skill.
  • Santa Fe deliberately mixes biologists, epidemiologists and physicists.
  • A talk on grid failure gave Gurley a lens on why the euro is too tightly coupled.

the biggest breakthroughs come from people that that had a different middle framework and move over and then see things differently

Bill Gurley · 1:51:00

the best solution is smaller communities that are Loosely coupled

Bill Gurley · 1:53:30
#book-recommendations#range#far-analogies#santa-fe-institute#cross-disciplinary

Story· 4

Story06:30

Passing on Google in 2002, and what it taught him about investing rules

Gurley walks through what he calls the biggest mistake of his career. Google presented to Benchmark with about 25 employees, and Benchmark did not chase. The externals looked bad — Yahoo was down from 80 to 10, Excite was going bankrupt, and two PhD founders insisted they could run the company. He is careful to note that John Doerr and Michael Moritz, the two best VCs of the era, both said yes, so no amount of study guaranteed the call. The lesson he draws is about rules themselves: things change, variables are dynamic, and the minute you set a very hard rule you may be setting yourself up for a mistake. The counterweight is asymmetry — you lose your money once, but a Google returns ten thousand times.

  • Benchmark saw Google at 25 employees and did not pursue the round.
  • Search looked dead externally, and PhD founders wanting to be CEO was a red flag.
  • Doerr and Moritz both said yes, so the miss was not simple negligence.
  • A partner asked Page what would close it; Page said 120 pre and it closed at 80.
  • Hard rules are dangerous in a system where every variable is dynamic.
  • Asymmetric payoffs force a bias toward the positive case.

let me work my way back to it because it's clearly the biggest mistake of my career

Bill Gurley · 07:00

the minute you set a very hard rule then you might be setting yourself up for a mistake

Bill Gurley · 08:30

what does it take to get this deal done right now

Bill Gurley · 10:30
#venture-capital#google#investing-rules#asymmetry#biggest-mistakes
Story25:00

The food analyst's framework that turned into a 100x call on Dell

All the analysts at Gurley's firm shared a floor, which is how he became close with Michael Mauboussin, then the food analyst. Mauboussin had just read the Stern Stewart and McKinsey material on return on invested capital and was proselytising it through the analyst group. Gurley, a sponge at the time, took the framework and ran it across all his companies. Dell's ROIC was roughly twenty to one against the industry — so extreme that Mauboussin initially did not believe it and they went back into the numbers. Simultaneously Dell had stumbled on a bad options trade and laptops catching fire, so the stock sat at six times earnings. Gurley went to a strong buy on a broken stock and it rose 100x, putting his name on the map with the buy side.

  • Shared floors put a tech analyst next to a food analyst.
  • ROIC came from Stern Stewart and McKinsey's Valuation, unknown in tech coverage.
  • Gurley ran it across his whole coverage universe rather than one company.
  • Dell's ROIC advantage was roughly 20 to 1 versus peers.
  • Mauboussin refused to believe the result until they re-derived it.
  • A broken stock at 6x earnings plus a hidden moat produced a 100x call.

I took the framework and ran it on all of my companies

Bill Gurley · 25:30

the first time I showed him he didn't Michael didn't believe it was true

Bill Gurley · 26:00

we went to a strong Buy on a broken stock and it went up 100x in the

Bill Gurley · 26:30
#roic#cross-disciplinary#dell#equity-research#contrarian-analysis
Story30:30

Brian Arthur to OpenTable to Uber: laying a network over a wasteful industry

Gurley traces network effects to Brian Arthur's 1996 HBR piece, Increasing Returns and the Two Worlds of Business, which he encountered through the Santa Fe Institute and calls the first real treatment of the idea. That thesis justified OpenTable despite an appalling rule set — selling PCs to restaurants that had no connectivity, forcing Benchmark to partner to get broadband installed. The bet was stated as a loop: enough restaurants pull consumers, and consumers force the remaining restaurants on. They built liquidity city by city starting in San Francisco. The proof came in a board meeting: the model required four restaurant closes per rep per month, the team was at 7.7, and one rep had closed 35 — the last rep in San Francisco, where penetration hit 90 percent. Gurley then deliberately reused the pattern to find Uber.

  • Brian Arthur's 1996 HBR article was Gurley's introduction to network effects.
  • Network businesses tend to winner-take-most, which is why they produce outliers.
  • OpenTable violated every deployment rule but the network thesis held.
  • Liquidity was built city by city, not everywhere at once.
  • One rep closing 35 in a month was the visible signature of 90 percent penetration.
  • Gurley reused the pattern to search for a network over a wasteful industry, finding black cars.

I was trying to think of other Industries where if you put a network on top of it it would absorb waste and make it…

Bill Gurley · 30:30

we built liquidity City by City

Bill Gurley · 34:00
#network-effects#marketplaces#opentable#uber#liquidity
Story1:34:00

Bezos, burner phones and the last person with good judgment

Gurley calls Bezos probably the best entrepreneur he has been around, and picks the least-discussed trait: Bezos built an organisational framework that runs the whole company on what he believes, without touching every decision. His evidence is a chance conversation with an Uber driver who had to be at a San Jose Amazon warehouse at 2:30, where Amazon loaded his car with packages, handed him burner phones and a manifest, and booked the trip over Uber. This was early same-day delivery, run by a company worth hundreds of billions as a hack on top of a competitor's platform. Gurley says no other large company would run that project — none, zero — and most firms past 20 or 30 million in revenue would refuse on accounting grounds. The stopping rule is the crux: an Amazon experiment dies when the last person with good judgment gives up.

  • Bezos institutionalised his beliefs into an organisational framework, not just decisions.
  • Amazon's early same-day delivery ran on burner phones and rides booked over Uber.
  • Most companies past 20 or 30 million revenue would veto that experiment.
  • Experiments are killed when the last person with good judgment gives up.
  • Big companies run one test and quit; startups cannot quit, so they get more shots on goal.
  • The six-page memo read at the start of meetings forces thinking a slide deck lets you skip.
  • AWS launching out of a consumer internet company is, to Gurley, a top-five business move ever.

this is a company that's worth 100 billions of dollars that is running an experiment on top of uber

Bill Gurley · 1:36:00

said when the last person with good judgment gives up

Bill Gurley · 1:37:30

forced to write a six page paper it's much harder to put that together than it is a five

Bill Gurley · 1:39:00
#bezos#experimentation#amazon#corporate-innovation#decision-frameworks

tactic· 2

tactic17:30

How Gurley made the All-America team: twenty interviews and a Palm Pilot full of names

The Institutional Investor ranking is a poll of the buy side, so Gurley worked backwards from the voters. Before starting the job he asked each salesperson for one client who would give a new analyst 30 to 45 minutes, and completed about twenty of those conversations with nothing to sell. The buy side told him they did not want buy or sell calls; they wanted analysis that made them think differently. He then attacked distribution, reasoning that a better-known analyst is more powerful and impactful even though the firm wants your content contained. He leveraged the sales force for fax numbers, then at Stewart Alsop's Agenda conference bought a Palm Pilot preloaded with every attendee's contact details at roughly 70 cents a name and mailed his weekly newsletter to the industry's top 400 or 500 people.

  • The II All-America ranking is literally a poll of buy-side clients.
  • Twenty pre-start interviews, framed as service rather than sales, revealed what clients valued.
  • Clients wanted differentiated analysis, not ratings changes.
  • A more widely known analyst is more powerful, so distribution is a separate axis from quality.
  • The Agenda conference sold Palm Pilots preloaded with every attendee's contacts.
  • Gurley ran the math at ~70 cents per influential name and bought the list.
  • The same publish-and-push loop migrated from fax to email to Twitter.

I just want to know how I can serve them best

Bill Gurley · 18:00

probably the most successful hack of my

Bill Gurley · 21:00

I spammed the four or 500 most important people in the tech industry with my Weekly Newsletter

Bill Gurley · 23:00
#customer-discovery#distribution#personal-brand#wall-street#growth-hacking
tactic38:30

Bet sizing, board seats as the real constraint, and Benchmark's equal partnership

Asked about check sizes, Gurley says Benchmark is essentially a market taker on the first cheque — the competitive dynamic defines what a Series A looks like, and they will not miss an outsized winner over price. What they care far more about is ownership percentage. The binding constraint is not capital at all: because Benchmark puts a partner on the board of every investment and usually becomes the largest shareholder there, board seats limit the number of deals. He then explains the structural precondition for their process. The founding partners created an equal partnership, breaking with the tenure hierarchies of private equity, law and real estate. Every new partner gets an equal seat and an equal share on arrival, which produces three emergent properties: every voice is heard, partners root for each other rather than compete, and judgement passes down through partner meetings.

  • Benchmark is a market taker on first cheque size, not a price optimiser.
  • Ownership percentage matters far more to them than the size of the cheque.
  • Board seats, not capital, are the real constraint on deal count.
  • The founding partners created venture's first equal partnership.
  • New partners get an equal seat and equal share from day one.
  • Equality makes voice, mutual support and candid disagreement emergent rather than enforced.
  • New partners bring every company they meet; the group debrief is the apprenticeship.

our limitation is our board seats more than the capital

Bill Gurley · 13:00

the founding partners of Benchmark did something that was not hadn't been done before in Venture which is they created an equal partnership

Bill Gurley · 40:30

here because we're going to divide it equally I want our new partners to be as successful as I possibly can

Bill Gurley · 42:00
#bet-sizing#ownership#partnership-structure#benchmark#incentive-design

debate· 4

debate35:00

The 17 percent CFO, McKinsey's 900,000 phones, and the wrong way to size a market

In 1999 Gurley recruited a public-company CFO into OpenTable, who shortly announced he was quitting because his model said the business would never work. The model froze penetration at 17 percent in every city, a number carried over from retail, where no business the CFO had worked with exceeded that share. Gurley's answer was that because he believed the network effects, they were going to 99. The CFO left. Gurley generalises with two more cases: McKinsey sizing the global mobile phone market at 900,000 units, and an NYU valuation professor capping Uber at two or three billion by taking the taxi market as the upper limit — which prompted Gurley's rebuttal, How to Miss By a Mile. His verdict is that TAM conservatism costs an investor more than it protects them.

  • OpenTable's CFO quit over a model with penetration frozen at 17 percent.
  • The 17 percent ceiling was a retail benchmark imported into a network business.
  • Gurley's counter was 99 percent penetration, and the business validated it.
  • McKinsey sized the global mobile phone market at 900,000 units.
  • Damodaran capped Uber at the taxi market; SF Uber was already 20x SF taxis.
  • A 10x better product grows the pie rather than dividing the existing one.

so we look at the model and I dive in and he has frozen penetration in each City at 17

Bill Gurley · 35:30

I wrote a reply called how to miss by a mile

Bill Gurley · 1:32:00
#tam#market-sizing#forecasting-errors#uber#network-effects
debate1:22:00

Twelve years on the Second Life board, and why Gurley thinks Meta's premise is wrong

Gurley read Snow Crash on release, thought it was the best thing he had ever consumed, and was knocking on Philip Rosedale's door when Second Life started. He served on that board for 12 years, giving him unusual first-hand data on immersive worlds. His conclusion separates gaming and escapism, which work, from the idea that people want to live ordinary real-world experiences virtually, which does not. Young people role-play; a handful of adults do too. But many of the most engaged users were seeking escape, sometimes from mental health difficulties or a hard patch — and both Snow Crash and Ready Player One were dystopias. They ran in-world board meetings, and Zoom is simply a better substitute, which is a Porter-style substitution argument. He estimates that shutting the VR effort would double Meta's stock.

  • Gurley served on the Second Life board for 12 years after reading Snow Crash.
  • Gaming and escapism work; virtualising ordinary real-world activities does not.
  • Heavy immersive users often skew toward those seeking escape.
  • Snow Crash and Ready Player One were both dystopian, which is the tell.
  • Zoom is a superior substitute for the in-world board meeting.
  • Meta spends five to ten billion a year on VR; Gurley thinks stopping doubles the stock.
  • Sunk-cost and pot-committed bias make a reversal unlikely.

Zoom is an amazing substitute which is one of the Frameworks from competitive strategy

Bill Gurley · 1:24:00

I think if they shut down the VR effort not only will the profitability would soar because they're spending real money like five to ten…

Bill Gurley · 1:25:00
#metaverse#second-life#substitution#sunk-cost#meta
debate1:54:30

Be less tribal: the billboard, and the intellectual-consistency test

Given a billboard for billions of people, Gurley writes: be less tribal. His claim is strong — he cannot imagine an activity that turns off more brain cells than tribal affiliation, and he rates political bias as more powerful than confirmation bias or sunk cost, despite all the Nobel-winning work on those. He has friends across the spectrum and insists the failure runs both ways. His examples are deliberately symmetric: gerrymandering is horrible and both sides do it; regulatory capture on one side means banks and big pharma, on the other credit unions, teachers' unions and police unions. If he were teaching it, he would start by showing people the documented cognitive biases, then push the consistency test — if a behaviour is unacceptable when the other side does it, name your actual principle.

  • Gurley's billboard message is simply: be less tribal.
  • He rates political bias above confirmation bias and sunk cost in strength.
  • The critique is symmetric by design, with examples from both sides.
  • Gerrymandering and regulatory capture are his paired illustrations.
  • A lecture would start with the documented catalogue of cognitive biases.
  • The core test is intellectual consistency: what is your actual principle?
  • He notes tribal people tend to view their own side as near perfect.

I would put be less tribal be less tribal

Bill Gurley · 1:55:00

I can't imagine an activity that turns off more brain cells than tribal affiliation

Bill Gurley · 1:55:00

why does anyone want to be intellectually inconsistent if you think it's an ends to a means then you're just in a fight

Bill Gurley · 1:57:30
#tribalism#cognitive-bias#intellectual-consistency#politics#critical-thinking
debate2:05:00

Regulatory capture: the TED talk he would give instead

Barred from venture, investing and career advice, Gurley says his TED talk would be regulatory capture, underpinned by a belief that capitalism and democracy will eventually destroy one another. The mechanism: in heavily regulated industries the incumbents end up writing the legislation, locking themselves in and building a moat — Porter again — through law rather than product. His flagship example is the 44 billion dollar programme paying doctors to implement EHR systems, designed with Epic's CEO on the healthcare advisory board, followed by another 20 billion for a meaningful-use phase paying people to actually use the software they were already paid to install. He points at pharma, banks and telcos as the most broken sectors, notes UK faster payments landed 17 years ago while ACH still takes three days, and describes being told to assemble 15 donors with maximum cheques just to meet a congressman.

  • Incumbents in regulated industries write the legislation and build legal moats.
  • The 44 billion EHR incentive programme was shaped with Epic's CEO on the advisory board.
  • A further 20 billion meaningful-use phase paid firms to use software they were paid to install.
  • Pharma, banks and telcos are Gurley's most-captured sectors.
  • UK faster payments arrived 17 years ago; US ACH still takes three days.
  • FedNow has sat on the books for a decade against bank and Visa lobbying.
  • Access required assembling 15 people with maximum donations, three separate times.

probably regulatory capture okay I have this core belief that capitalism and democracy who eventually destroy one another

Bill Gurley · 2:05:00

the incumbents typically end up being the ones that write the legislation and typically lock themselves in build competitive Advantage build a moat back to…

Bill Gurley · 2:05:30

they all need to bring the maximum check that they can donate

Bill Gurley · 2:09:00
#regulatory-capture#healthcare#lobbying#payments#policy