Pat Grady - Relentless Application of Force - [Invest Like the Best, EP.378]
What this covers
The conversation with Pat Grady moves between three subjects: how Sequoia's growth team picks companies, how the firm is built to keep its edge, and what separates enduring companies from the rest. On the firm's internal design, he describes a Flattened Power Structure of anonymous initial votes and senior partners speaking last. He tells how Doug Leone rewrote the legal documents in 2012 so he could be voted out like anyone else. He also argues that being demanding of one another is itself a form of support, and he explains why Sequoia concentrates knowledge in a few senior investors while growing a large operator-staffed platform.
The investing mechanics get the most room. Grady describes assessing founders through long walks and questions about childhood, and treating people as vectors, so that output is a Vector Sum of Individuals. He explains the Emerging Market Leader Criterion, the habit of underwriting to a potential 10x, and how unique and compelling value propositions should appear in gross margin and operating margin. Two case stories anchor this: Snowflake's alarming early numbers, which led Sequoia to raise its stake from $50 million to $200 million, and Michael Moritz in Stockholm diligencing Klarna. His claim that venture capital is not an asset class is the most openly contested position here.
The AI section is the most forward-looking. Grady puts current LLMs about four orders of magnitude behind the brain in energy efficiency, suggests optimization alone would transform industries even if capabilities froze, and compares ChatGPT to Netscape. The closing stretch is more personal. He recalls nearly being fired early at Sequoia until Jim Goetz stepped in, and ends on "relentless application of force" as the mark of legendary companies.
Pat Grady argues that Sequoia's growth-investing edge comes from judging founder character and where the market is heading rather than from financial models, and that a culture of flattened power, clarity-driven communication, and relentless effort sustains outperformance.
- Founder character and market trajectory are the two things that must be right; financial models extrapolate poorly for outlier companies
- Power dynamics are flattened (anonymous votes, senior partners speak last) so influence follows expertise, not tenure
- Winning companies are judged on emerging market leadership, unique and compelling value propositions, and a team-driven sustainable advantage
This asset isn't compiled yet
You're seeing its claims, ranked. Compile it to build the argument threads, weight them, and check each claim against your library — the full view.
Venture capital is not an asset class because less than 1% of startups account for about 99% of the market cap created, so buying an index of startups yields mostly noise and poor funds, and it attracts participants motivated by fame, exits, or fees rather than serving founders.
“venture capital is not an asset class. What he means by that is less than 1% of the companies that get started end up accounting for 99% of the market cap that's created.”
State-of-the-art LLMs are roughly four orders of magnitude less efficient than the human brain in converting energy to computation, which implies a better, far more energy-efficient architecture exists and will likely be found long before AGI.
“state-of-the-art LLMs are about four orders of magnitude less efficient than your brain.”
If frontier model capabilities froze at today's level, optimization work alone to make models cheaper, faster, and easier would revolutionize almost every industry, because current capabilities are powerful and have only begun to be harnessed.
“if you froze capabilities today, and the only thing that you invested in was optimization, making it cheaper, making it faster, making it easier, if you did that, you would revolutionize almost every industry on earth.”
Neural nets have existed for 70 to 80 years but only became practical once compute, bandwidth, data, and talent converged, and ChatGPT in fall 2022 is likely this generation's Netscape moment that gave everyone a visceral sense of what LLMs could do.
“we think will end up being this generation's Netscape moment.”
A unique value proposition should show up in gross margin because a truly unique product makes the company a price setter, while a compelling value proposition should show up in operating margin because it needs little sales and marketing to win customers, with signals such as organic acquisition, LTV to CAC, or payback period.
“if your product is truly unique, you should be a price setter, not a price taker.”
Sustainable competitive advantage is more dynamic than a moat because it is built every day, and the top advantage Sequoia sees at companies is not network effects, ecosystem, or IT but the DNA of the team.
“the #1 sustainable competitive advantage that we see at the companies, it's not a network effect. It's not an ecosystem advantage. It is the DNA of the team.”
Most people assume demanding and supportive are inherently in trade-off, but Sequoia's view is that being demanding of one another is itself supportive, and performance comes from demanding excellence first of oneself and then of others.
“Most people think there's an inherent trade-off between these two things, we do not. We think that being demanding of one another is being supportive.”
Emerging market leader means investing in the company Sequoia believes will be the market leader tomorrow, not the one leading today, because the number one player captures a disproportionate share of market cap, so investing in a number two or three yields limited outsized returns.
“whoever ends up #1 in the market doesn't just have their proportional share of the market cap. They have a disproportionate share of the market cap.”
To assess a founder's character beyond LinkedIn, Pat goes on long walks and asks about childhood, parents' characteristics, siblings, happiest and biggest childhood mistakes, because deeper answers reveal what drives them and whether they will build for the long term.
“The way that I like to do that is to go for a long walk with founders and try to understand who they are in all the ways that don't show up on LinkedIn.”
When Doug Leone became senior steward in 2012 and discovered he alone could not be removed, he changed the legal documents so he could be voted off like anyone else, because being safe breeds complacency that leads to mediocrity and decline.
“as soon as you're safe, you start to go a little bit complacent and as soon as you get complacent, you start to become mediocrity, and then you're on the long path to inevitable decline.”
Venture partners who have been around and had successes tend to become deferential, but each investment is unique, so Sequoia flattens power dynamics through anonymous initial votes and having senior partners speak last so influence is a function of expertise, not tenure.
“influence is a function of expertise, not a function of tenure or power or some other corrosive dynamic.”
Sequoia underwrites to a potential 10x-plus return, because experience shows an investment underwritten to 3x ends up as about 2x, while one that could plausibly return 10x may end up at 3x to 7x, and if the 10x upside is not visible the company is probably not good enough.
“if you underwrite to a 3x return, you end up with a 2x.”
Sequoia looks for a clear and internally consistent founder story supported by evidence rather than perfection, and investors should keep asking questions until the picture is clear, since without clarity they cannot understand the risks and expected return.
“The thing that we're looking for is not perfection. The thing that we're looking for is clarity.”
Reference checks should assess a person as a vector with magnitude, meaning how top-graded they have been at each life stage, and direction, meaning whether they care about and excel at what matters to them, because organizational output is the vector sum of its individuals.
“the output of an organization is the vector sum of its individuals.”
Sequoia's platform grew from 14 investors and 2 operators to 27 investors and about 65 operators, amplifying investor efforts with a homegrown CRM and data science system that holds more information on unmet companies than Sequoia had at final decision 15 years earlier, while creating compounding proprietary advantages.
“The information that we have available in that system for a company that we've never met is more than the information we would have had on the same company 15 years ago”
Because the only two things requiring humans in investing are building founder relationships and making the decision, Sequoia chose to concentrate the partnership's knowledge in few senior investors rather than dispersing it across hundreds, believing this lets each investor grow exceptional and attract outlier founders.
“there are only two things that you need a human being to do in the world of investing.”
During 2010 diligence on Klarna in Stockholm, Michael Moritz said the question was whether they could reach a few hundred million dollars of net income and that it would hinge on the strength of the engineering team, which proved correct because complexity of e-commerce integrations required elegant product-driven execution.
“The question is whether they can get to a few hundred million of net income, and the answer will come down to the strength of the engineering team.”
Sarah's firm Conviction Partners treats smallness as a weapon, keeping funds, teams, and portfolios smaller than possible, optimizing for quality, because quality attracts growth while chasing growth tends to produce mediocre companies.
“she believes that being small as a weapon and that you don't get advantages out of scale in the venture capital business, you get advantages out of quality.”
Pat nearly got fired early at Sequoia after misunderstanding his role as teaching Sequoia to invest like Summit, when five of six general partners wanted him let go, and Jim Goetz intervened because he saw Pat desperately wanted to do the right thing.
“five of six general partners said that I should be let go, and Jim Goetz was the one who said over my dead body.”
Founder character cannot be known with certainty, but it can be assessed with better odds, and at the growth stage the real difference between Sequoia and other investors comes from understanding the people, not from better cohort analysis or customer calls.
“I think it's going to be a function of actually understanding the people, and when that model ends at the out year five years from now, has that founder gotten 10 times better”
Markets do not know how to assess founder quality because human beings are harder to decompose than product architecture or income statements, and most investors lack the language and courage to ask the personal questions required, so founder quality is often mispriced.
“I don't think markets know how to assess founder quality. It is comparatively easy to decompose the architecture of our products or the nuances of an income statement than it is to decompose a human being.”
Six months after investing in Snowflake, portfolio numbers looked bad on revenue, gross margin, and free cash flow, but investigating with the FP&A VP showed each shortfall had a positive cause such as larger deals ramping slowly, global availability zones, and higher rep productivity, leading Sequoia to expand its investment from $50 million to $200 million.
“all of the reason that the numbers were bad turned out to be good reasons.”
What separates truly legendary companies from the rest is the relentless application of force, meaning sustained effort driven by a durable motivation such as mission, craft, or responsibility to employees rather than fame or unicorn status.
“the thing that probably separates the legendary from the rest is, in fact, the relentless application of force.”
The biggest mistake investors make is selling by telling founders how awesome the firm is, when founders care about how awesome they can become, so Sequoia tries to understand the founder's vision and show it believes in that vision before discussing resources.
“I think the biggest mistake people make is selling by telling founders how awesome you are. Founders don't care how awesome you are. They want to know how awesome they have a chance to become.”
Pat's three criteria form a sequence where emerging market leadership signals revenue scale, unique and compelling value proposition signals margin structure, and sustainable competitive advantage protects those margins, which together determine the out-year financial model rather than linear extrapolation.
“if you have an emerging market leader, chances are you'll have good revenue scale.”
When evaluating AI products, Sequoia spends less time underwriting architecture and more time underwriting the customer problem, uniqueness, durability, and alternatives, because the technical approach only matters to the extent it creates unique and compelling value, and investors who overweight their architectural opinions make a mistake.
“the technical approach only matters to the extent it does something unique and compelling for the customer.”
Killer instinct comes from entering a mode where only the objective function is visible, and Doug Leone lives this way such that personal discomfort and risk do not enter his calculation, reflecting ultimate humility and mission orientation.
“You go into a mode where you are purely driven by the objective function”
Harvey is a legal assistant that performs the work of a first-year big-law associate, and partners A/B tested it against associates and found it just as good, turning a six-hour task into six seconds, with the long-term ambition of democratizing access to legal services.
“the test that might have taken six hours instead takes six seconds.”
Sequoia treats its business as an apprenticeship business, and Pat argues that judgment and values are learned from environment, so founders' lack of good early mentors explains differences in their behavior and that investors should help uncover the goodness beneath rough exteriors.
“I think people are a function of their environments and your environments teach you certain lessons”
A standard 2012 Airbnb growth memo modeled a 2.5x return using linear extrapolation, but a follow-up memo argued it could become a $100 billion company and return the fund several times, and the conviction to make that claim is what moved Sequoia to invest.
“Page 1 of the memo tends to become a $100 billion company and here is why.”
Forrester estimated a $150 million TAM for cloud identity when Okta went public in 2017, yet Okta already had about that much revenue, which shows that static market-size snapshots underestimate markets and that revenue projections should come from market dynamics and company position rather than the financial model.
“the TAM for cloud identity was $150 million. Well, Okta at that time, had, I think, a little more than $150 million of revenue.”
Market determines how big a company can get and people determine how big it will get, so a physics-style view of system rules is complementary to a human view because individual agents operating within the system determine outcomes, and combining both yields the best decisions.
“The market determines how big the company can get and the people determine how big the company will get.”
The all-time great investers share two superpowers, understanding human beings and having a sense for where the world is going, and these two things are what anyone taking a long-term equity stake must get right across public, private equity, and venture.
“Number one, understanding human beings. Number two, having a sense for where the world is going.”
Doug Leone questioned founders intensely for 30 to 40 minutes, then summarized what he heard, naming both the vision and the business flaws, which made about 95 percent of founders eager to work with Sequoia.
“Maybe not 100% of the time, but 95% of the time after Doug unfurled one of those, the founder would be desperate to work with us”
Early in his career Pat opened meetings with older founders by saying he didn't know their market but saw a chance for them, using research on their website and customers, which validated their ambition and earned the benefit of the doubt, and Pat still believes making founders feel seen is the most powerful move.
“make a founder feel like they're seen, you understand them and try to validate their ambition”
Importing a growth equity playbook from Summit Partners into Sequoia failed because a firm must first understand its own identity and then extend that into growth investing, rather than teaching the firm to invest like another firm.
“I misunderstood my job as, teach Sequoia how to invest like Summit, and the thing that I should have done was understand who Sequoia is and then extend the Sequoia into growth.”
A great investment memo is about 3 pages, with the exhibits in an appendix, and it should convey an accurate point of view on what is good, what is bad, and the so-what, because the goal is clarity and understanding rather than impressing people with intelligence.
“A great investment memo is 3 pages, not 12 pages.”
If returns to scale are diminishing, the AI industry could pivot to a small team of geniuses designing new architectures, shifting compute from training toward inference for planning and reasoning, and aggressive optimization, creating a cash-generative business with a dominant position but with the risk of not having reached diminishing returns.
“we're going to move some of the compute from training into inference.”
In 1999-2000 the smart money favored eBay's marketplace model over selling books online, but Amazon won because Jeff Bezos and the culture he created compounded advantages over time, illustrating that founder DNA beats seemingly superior business models.
“It turns out that one of those companies had Jeff Bezos and the other one didn't.”
On each investment there are two roles: the sponsor secures the investment and makes the internal case, while the wing person supports the sponsor by doing the modeling, memo, and customer calls, and being a good teammate means doing whatever job the mission needs.
“There's the sponsor and there's the wing person. And sponsor's job is to secure the investment and to make the case internally.”
A firm that adds other offerings to its core, like a chicken restaurant adding burgers, dilutes its identity, so a venture firm known for one focus such as early-stage AI attracts the right founders, and it is advice that also applies to new software companies.
“Most companies have a chicken issue where they want to do the chicken and the burger and just focusing on the chicken is important.”
Investors should operate at the right level of ambition and focus on the first-order issue, as Moritz did by reframing a deal around a few hundred million in net income and engineering team strength instead of a long list of secondary questions.
“you got to zoom out and make sure that you're operating at the right level of ambition”
Sometimes leadership means doing the work visibly so others learn how it should be done, which in an apprenticeship business is more helpful than one-on-ones, mentorship, or structured feedback.
“sometimes leadership actually just means doing the work so that people can see how it's supposed to be done”
Don Valentine did not ask for any money when the first generational transition happened in the mid-1990s, which imbued the partnership with a sense of stewardship that is a big part of what drives Sequoia's performance drive.
“Don didn't ask for any money whatsoever. He just handed it over. In doing so, he imbued the partnership with this sense of stewardship, which is actually a big part of what drives us.”
Sequoia partner Alfred Lin decided to back DoorDash after a dinner with Tony Xu before the Series A, where Xu's grasp of business fundamentals was so deep that Alfred concluded the investment was about the founder's ability to create compounding advantages rather than the business model or market.
“we have to make this investment not because of the business model, not because of the market, but just because of the founder.”
Sequoia's objective function is to maximize net multiple of money returns for its limited partners, and founder NPS is a second metric, with the two being the metrics it cares about most, because founder referrals are what carry deals over the line.
“Our objective function is to maximize net multiple money returns for our limited partners, not to maximize founder NPS”
The kindest thing anyone has done for Grady was give him a chance, from parents and scholarship mentors to John Carroll at Summit and Doug Leone at Sequoia, and Jim Goetz defended him when five of six general partners wanted him fired.
“at one point, apparently, five of six general partners said that I should be let go, and Jim Goetz was the one who said over my dead body”
Harvey aims to use AI to make world-class legal services affordable for the broad market, so that high-end firms use it as an assistant while everyone else gets it as a service, expanding rather than replacing human lawyers' market.
“it turns out that with AI, in the fullness of time, we can provide world-class legal services, and we can do it at a tiny fraction of the price.”
Grady's competitive intensity can override social awareness, as when he yelled at a founder during an axe-throwing contest to keep throwing, showing that a pure objective-function focus crowds out other considerations.
“I had lost track of the social graces of yelling at somebody to put their arms down and throw more axes because the objective function was the only thing that I could see.”
Stable Diffusion in summer 2022 moved AI from researchers to researchers plus ML engineers, and ChatGPT that fall moved it to all engineers, product managers, founders, consumers, and Fortune 500 boardrooms, with each step accelerating application development.
“That took the AI market from researchers to researchers plus machine learning engineers.”
The term leptokurtic return profile was introduced by Roelof Botha, who Grady says may be the only licensed actuary active in venture capital.
“That comes from Roelof, so Roelof might be the only licensed actuary who is active in the venture capital business today.”
Sequoia invested in Airbnb at the seed stage in 2009, a few weeks before the market bottom, when the idea seemed crazy.
“a few weeks before the market bottomed out, we got into business with Airbnb at the seed stage. That seemed like a crazy idea”
Pat hopes people would say at his retirement that he was a top performer and a top teammate, reflecting the Boston College motto ever to excel and Jesuit men and women for others, which map onto Sequoia's two core values of performance and teamwork.
“I think that the thing that I would hope to hear is that I was a top performer, but also a top teammate.”
The industry would improve if the share of participants with impure motives, such as founders seeking fame or investors seeking fees and publicity, were reduced in favor of founders focused on customers and investors focused on founders.
“I would try to slice off some of the people who are participating without the most pure motives”
Grady and his wife simplified their family values into four principles, work hard, be kind, think for yourself, and family first, which they believe can derive the right answer to most situations and explain misbehavior.
“they're just work hard, be kind, think for yourself, family first. That's it.”
Magnitude in a reference check is assessed by asking whether the person was the best at each stage of life, such as high school, college, and each job, though this is a blunt measure that must be combined with direction.
“How good has this person been at each step of their life? Were they the best person in their high school?”
Sequoia had each investment team member write a 2030 vision memo, removed the names, and voted, so the best ideas would win rather than the ideas of whoever had been around longest.
“each person on our investment team wrote a memo that was their vision for Sequoia in 2030. Then we took the names off of all the memos and voted on it”
Doug Leone's transparent negotiating style, where he stated what Sequoia would pay and guessed the counterparty's price and met in the middle, taught Pat that the simpler and more transparent you are, the more people trust you, and trust makes life easier.
“the simpler you can keep things, the more straightforward you can be, the more transparent you can be, the more people are going to trust you”
The kindest thing anyone did for Pat was to see past his prickly, rough-around-the-edges early self and help his positive attributes flourish, which is also how Sequoia treats founders.
“to see through that and to see whatever goodness or whatever positive attributes might have been hiding inside and to help those things come out and flourish over time.”
Sequoia believes anybody can beat it on any given day because competitors are smart, hardworking, and have killer instincts, so the Sequoia brand alone never wins a competitive deal and every pitch must be at the team's absolute best.
“we can't just waltz in and hope that the Sequoia business card is going to give us the advantage.”
Clear family and team values make decision-making easy because they can be used to derive the right answer to most questions, and Sequoia's growth team values (aggressive but humble, demanding and supportive, I give a shit, zero bull, strong under scrutiny) collectively produce most effective long-term behaviors.
“when your values are clear, decision-making is easy.”
Because Pat is structured and linear rather than creative, he builds scaffolding for himself through long-term plans and annual OKRs that cascade into quarterly and weekly goals, which creates the pressure he needs to perform.
“I erect scaffolding around myself that attempts to create that pressure for me.”
Sequoia partners put far more pressure on themselves than on each other, and they want that self-imposed pressure because it helps guarantee, or at least improve, performance.
“yes, we try to put pressure on each other, but we put far more pressure on ourselves, and we want that pressure because we know that, that pressure is the thing that guarantees performance.”
Pat admires extroverted, charming people who light up rooms, and if he could change one thing about himself he would become more extroverted and charming, though he has worked at networking without finding it comfortable.
“that's probably the magic wand is to make myself a little more extroverted”
By around 2017, after roughly a decade focused on cloud (2007-2017) and with mobile app store rankings stagnant, Sequoia began looking for the next platform shift and chose data as the shorthand because the best application experiences were fueled by machine learning on data, which led to Snowflake, Confluent, dbt, and NLP investments.
“it felt like we were getting fairly late cycle as far as these technology platform shifts go.”
Early in his career Pat relied on brute-force meeting volume and spent hours crafting single emails to get in front of founders, and that effort usually worked, opening access that Doug Leone then converted.
“I would spend hours writing an email, many hours to write a single email to get it exactly right”
The proposed efficiency metric for an LLM is math per Cheeto, where energy input is a bag of Cheetos and computation output is math, and humans currently do far more math per Cheeto than the best LLM.
“the metric for the efficiency of an LLM should be maths per Cheeto.”
When a high-trust relationship exists before the deal, the decision is fairly straightforward as long as the offer is not offensively different from market terms.
“As long as the offer that we make is not offensively different than market, it's a fairly straightforward decision”
Sequoia selects for a drive, often rooted in a chip on the shoulder or personal demons, during interviews because that internal motivation predicts the performance they seek.
“Whatever caused the chip on their shoulder or whatever the demons that drive them might be, it is something that we try to select for in the interview process”
When Pat joined Sequoia at 24 as its youngest hire, he berated himself on daily commutes and spent 48 hours on his self-review, which was like self-therapy, though he never fully broke.
“I probably spent 48 hours on my self-review, and it was almost like a self-therapy session”