← Back to list
Colossus (Invest Like the Best / Business Breakdowns)Podcast18 Jun 2024Source: joincolossus.comHost: Patrick O'Shaughnessy

Pat Grady - Relentless Application of Force - [Invest Like the Best, EP.378]

In plain words

This piece covers Sequoia investor Pat Grady's philosophy. He says a company's long-term edge isn't tech or market share but its team's DNA—Amazon beat eBay because of Jeff Bezos, not its business model. He's bullish on AI, believing even current models can transform industries through optimization. Key holdings: Snowflake (he added investment from $15M to $200M), Zoom (invested when revenue was ~$85M with 80%+ margins), and Harvey (an AI legal assistant that turns 6-hour tasks into 6 seconds).

AI SummaryAI-generated · may contain errors · verify against the original

At a Glance Sequoia Capital's long-term growth investor Pat Grady shares his investment philosophy and core framework on the program. He proposes three key criteria for evaluating companies: market potential, product differentiation, and team execution, emphasizing that the "Relentless Application o

~20 min full read · 11 sections
Deep Analysis

Pat Grady - Relentless Application of Force - [Invest Like the Best, EP.378]

At a Glance

Pat Grady is a long-term growth investor and senior partner at Sequoia Capital, having participated in iconic investments including Snowflake, Zoom, ServiceNow, Qualtrics, Okta, HubSpot, Notion, and OpenAI. The main thread of this episode: how Sequoia's internal pressure culture shapes investment decisions, and Pat Grady's three criteria for evaluating companies (market potential, product differentiation, and team execution). The most weighty judgment in the entire episode: Pat Grady believes the core of "sustainable competitive advantage" is not network effects or IP, but team DNA—"Amazon beat eBay not because of its business model, but because of Jeff Bezos"; DoorDash prevailed not because of its delivery model, but because of Tony Xu.


I. Sequoia’s Internal Culture of Pressure: The Legacy from Don Valentine to Doug Leone

Pat Grady believes that the root of Sequoia’s "healthy peer pressure" culture lies in Don Valentine’s act of "taking nothing" during the handover in the 1990s—instilling in partners a sense of "stewardship" toward the firm.

  • Historical Context: In the mid-1990s, Don Valentine transferred partnership rights to Michael Moritz and Doug Leone without demanding any buyout payment. Grady explains: "We were given a wonderful gift—Sequoia. The only requirement was to ensure that when we leave, it is better than when we took over."
  • Mechanism Breakdown: After Doug Leone became the top manager in 2012, he discovered that the legal documents stipulated only he was "safe" (could not be voted out). He immediately amended the documents to make himself subject to removal by vote. Grady paraphrases Doug’s logic: "Once you feel safe, you become complacent; once complacent, you drift toward mediocrity; then comes inevitable decline."
  • Specific Practices: Sequoia "flattens" partnership hierarchy through multiple mechanisms—investment votes are anonymous; senior partners speak last to avoid influencing discussions; in a strategic discussion five years ago, all partners anonymously submitted "Sequoia 2030 Vision" memos and voted anonymously. Grady says: "Influence should come from expertise, not seniority or power."

On a Personal Level for Pat Grady: He builds his own "scaffolding"—long-term plans → annual OKRs → quarterly OKRs → weekly/daily tasks. He recalls that when he first joined in 2007 (at age 24, the youngest employee in Sequoia’s history), he would spend the 30-40 minute drive home from work every day blaming himself for "not doing enough." His annual self-assessment takes 48 hours, "almost like self-therapy, constantly tearing myself apart."


II. Three Investment Evaluation Criteria: Market, Product, and Moat

Pat Grady outlines the three investment criteria that Sequoia's growth team finalized during two closed-door days at Pelican Hill, each corresponding to a financial dimension.

1. Emerging Market Leader → Revenue Scale

Grady emphasizes: "We are not investing in today's market leaders, but in companies we believe will become market leaders tomorrow."

  • Dual meaning: The market itself is "emerging" (today's TAM may be small, but there is reason to believe it will expand); the company is "emerging" to become the leader of that market.
  • Data support: When Okta went public in 2017, a Forrester report estimated the cloud identity TAM at only $150 million, while Okta's revenue at the time already exceeded that figure. Grady notes: "People find it hard to extrapolate from a static point in time to 5-10 years out. Ultimately, what determines long-term revenue projections is not the financial model, but market dynamics and company positioning."
  • Key logic: In the tech sector, the market leader's market cap is not distributed proportionally but "disproportionately" captures the majority of value. Investing in the second or third player may yield some returns, but it will not generate outsized returns.

2. Unique & Compelling Value Proposition → Margin Structure

Grady breaks it down: "Unique" corresponds to gross margin—if you have a unique product, you are a price setter rather than a price taker, and should achieve high gross margins; "Compelling" corresponds to operating margin—if the product is truly compelling, you do not need to "hammer" customers with sales and marketing spend.

  • Quantitative metrics: Figures such as new ARR ÷ sales and marketing spend, LTV/CAC, payback period, or "99% of new customers come from organic growth" should demonstrate the "compelling" nature of the value proposition.
  • Logic chain: Unique → high gross margin; Compelling → low sales expense → high operating margin. The combination determines the long-term financial model.

3. Sustainable Competitive Advantage → Moat

Grady deliberately distinguishes "sustainable competitive advantage" from "moat": "A moat implies it is already built and will protect you forever; a sustainable competitive advantage is an advantage that is built every day."

  • Core thesis: The number one sustainable competitive advantage is not network effects, ecosystem advantages, or IP, but team DNA.
  • Historical analogy: In 1999-2000, smart money bet on eBay (elegant marketplace model, strong defensibility), but should have bet on Amazon (selling books is a commodity business). "One company had Jeff Bezos, the other did not. It is the founder and culture that create compounding advantages over time."
  • Modern case: DoorDash faced numerous competitors, but "there is only one Tony Xu." Sequoia partner Alfred Lin met Tony at the seed round, but it was not until a dinner before the Series A that he was "blown away" by Tony's grasp of business details, and returned saying they had to invest—not because of the business model or market, but because of the founder.

3. Evaluating Founders: Understanding the "Vector of the Person"

Pat Grady believes the market underestimates the importance of founder quality in the right-tail distribution, because "deconstructing the architecture of a person" is far harder than deconstructing a product architecture or a profit and loss statement.

  • Methodology: Grady prefers to take "long walks" with founders, asking questions that cannot be found on LinkedIn—childhood experiences, traits inherited from each parent, relationships with siblings, the happiest moments of childhood, and the biggest mistakes. "These questions themselves may not reveal much, but as you dig deeper, you begin to truly understand who they are and what they value."
  • Goal of Reference Checks: Grady borrows Elon Musk's "vector" framework—the output of an organization is the sum of individual vectors. Vectors have both magnitude and direction. Reference checks aim to determine: Was this person the "best" at every step? Did they excel at what they cared about? This defines the vector's direction and magnitude.
  • Why the Market Fails at This: "Most investors don't know the language of deconstructing a person and are afraid to ask overly personal questions. Discussing a person at a partner meeting is difficult because the data points are opinions, not facts, and can easily trigger emotional conflicts. But ultimately, this is the most important factor."

Grady's Self-Identified Weakness: When asked what he would change about himself with a magic wand, he replied that he wished he were more outgoing, more charismatic, and better able to "light up a room." He envies the ability of former partner Carl Eschenbach and his wife Anna to "walk into a room and naturally command attention." "I might hide in a corner, hoping someone I know will come over and talk to me. I force myself into awkward social situations countless times, and I still feel uncomfortable."


4. From "Bad Numbers" to Good Investments: The Snowflake Case and "Keep Asking Until It's Clear"

Pat Grady uses the Snowflake case to illustrate: behind bad numbers may lie a good story; the key is to "keep asking until the picture becomes clear."

  • Background: Sequoia first invested in Snowflake when it had approximately $50 million in ARR. Six months later, the semi-annual portfolio review showed: revenue trailing the plan, gross margin below expectations, and free cash flow lower than forecast. Grady thought at the time: "Oh no, maybe we shouldn't have invested."
  • Truth: After in-depth discussions with Snowflake's FP&A head, Brad Floring, Grady discovered:
  • Revenue was trailing because "much larger-than-expected deals were closed, and these large deals take longer to recognize as revenue."
  • Gross margin declined because "globalization happened faster than anticipated, requiring availability zones to be opened in low-utilization regions; at the same time, full-scale Teradata replacements were requested, necessitating professional services (included in COGS)."
  • Operating margin declined because "productivity per employee far exceeded expectations, prompting the company to accelerate hiring of the sales team."
  • Outcome: All the "bad numbers" were actually good news. Sequoia increased its investment from an initial $15 million to $200 million. "You keep asking questions until the picture in your mind becomes clear. It doesn't need to be perfect, but it must be clear. If it's not clear, you cannot understand the risks you are taking or the returns you expect."

Grady's investment philosophy: "What we seek is not perfection, but clarity. The story must be internally consistent, the available evidence must support the story, and there must be no contradictions on important aspects."


5. AI Frontier: Market Judgments, Architecture Efficiency Gap, and the Harvey Case

Market Judgments

Pat Grady believes that even freezing current model capabilities (GPT-4o, Claude 3/4), optimization alone (cheaper, faster, easier) could "revolutionize nearly every industry."

  • Historical Context: Sequoia began focusing on "data" as the next platform shift around 2017—observing that the best application experiences were driven by machine learning. This led them to invest in Snowflake, Confluent, dbt, Hugging Face, OpenAI, and others.
  • ChatGPT's "Netscape Moment": In the summer of 2022, Stable Diffusion expanded AI from "researchers" to "ML engineers"; in the fall of 2022, ChatGPT further extended it to "all engineers, product managers, founders, consumers, and Fortune 500 boards."
  • Architecture Efficiency Gap: Grady cites observations from John Carmack and Andrej Karpathy—current state-of-the-art LLMs are about 4-6 orders of magnitude less efficient than the human brain (in terms of energy input per computational output). "Nature has proven that a better architecture exists. Before we reach widely recognized AGI, a different, more energy-efficient underlying architecture will inevitably emerge."

The Harvey Case: AI Legal Assistant

Grady considers Harvey "one of the first and best application companies representing the AI platform shift."

  • Founder Fit: Winston (legal background) + Gabe (AI research background)—"understanding both the problem and the solution simultaneously."
  • Product Status: Harvey currently serves as a legal assistant, capable of performing the work of a first-year associate at a large law firm. In A/B tests comparing Harvey with human assistants, partners found "Harvey is equally good and instantaneous." Tasks that originally took 6 hours now take 6 seconds.
  • Long-Term Vision: Not to replace lawyers, but to "democratize law"—currently, legal services are a "luxury product for the wealthy." High-end law firms use Harvey as an assistant; for the rest of the world, Harvey itself is the service.
  • Investment Evaluation Approach: Grady emphasizes "starting from the customer, not from the technology"—"Many investors place too much weight on their personal opinions about product architecture. Technical approaches only matter when they create unique and compelling value for the customer."

VI. Sequoia’s Platform Strategy and the “Chicken Filet Problem”

Pat Grady explains why Sequoia chose a “small but elite” investment team plus a large platform team, rather than expanding the investment team size.

  • Data: In 2007, Sequoia had 14 investment team members + 2 front-office operations staff (1 in talent, 1 in marketing). Today: 27 investment team members + approximately 65 front-office operations staff (marketing, talent, engineering, product, data science, design, client partnerships, etc.).
  • Two key advantages:

1. Amplifying investment team efficiency: With a proprietary CRM and data science system, the firm now has more information on companies it has “never met” than it had when making final investment decisions 15 years ago.

2. Compounding advantages: Historically, the only compounding advantages in VC were brand, culture, and networks—all relatively transient. What the platform team builds can compound. For example, the talent database has accumulated proprietary signals on hundreds of thousands of individuals.

  • Strategic choice: Sequoia decided not to expand the investment team to 270 people because “there are only two things that must be done by people—building relationships with founders and making decisions.” If knowledge and experience are dispersed across hundreds, no one stands out; if concentrated in the smallest possible group, everyone has the chance to become exceptional. “The person who hunts outlier founders must themselves be an outlier.”

The “Chicken Filet Problem”: Grady cites Jeff Richards of GCV—Chick-fil-A could attract more customers by adding hamburgers, but “you are known for chicken. People who want hamburgers should go elsewhere.” Most companies face the “chicken filet problem”—they want to do both chicken and hamburgers. Grady believes that Conviction Partners, founded by his wife Sarah, got it right: focusing solely on “early-stage AI companies” (Series A and earlier), optimizing for quality rather than scale.


VII. "Relentless Application of Force": The Defining Trait of Legendary Companies

Pat Grady argues that the core differentiator between legendary companies and the rest is the "Relentless Application of Force."

  • Source: A remark made by former Sequoia partner Michael Moritz to Faire founder and CEO Max Rhodes. When asked, "What keeps you going when your friends are out having fun and you're exhausted and want to give up?" Max replied: "It's Michael Moritz's voice echoing in my head—relentless application of force."
  • Deeper Question: What drives you to apply relentless force? Grady believes this touches your core—what you truly care about and why you are building this company. "If your motivation is to issue a press release announcing you're a unicorn, that may not be enough to sustain you through tough times. Motivation can be an obsession with customer problems, a passion for building, a pursuit of craftsmanship, or a sense of responsibility to hundreds of employees."

Grady's Critique of the Industry: He quotes partner Roloof Bota—"Venture capital is not an asset class." Less than 1% of companies create 99% of market value. If you treat it as an asset class and buy an index, you'll be drowned out by noise. "Most companies don't need to exist. They aren't solving important problems, or they aren't solving them in a unique and compelling way." He hopes to "cut out some participants with impure motives" and increase the number of founders who truly care about customers and investors who truly care about founders.


Mentioned Positions

Position Analyst Stance Key Data
Snowflake Bullish (added to position) ~$50M ARR at initial investment; initial $15M → added $200M; gross margin ~50% (at the time)
Zoom Bullish ~$85-90M revenue at time of investment; 80%+ gross margin; "wanted to burn cash but customer money came too fast"
Airbnb Bullish (seed round) Seed round investment in 2009; during 2012 growth round, a memo stated "has the potential to become a $100B company"
Okta Bullish Forrester report at 2017 IPO pegged cloud identity TAM at only $150M
DoorDash Bullish (Series A) Sequoia partner Alfred Lin pushed for the investment due to Tony Xu's "grasp of business details"
Klarna Bullish Invested in April 2010; Michael Moritz's key judgment was "whether it can reach hundreds of millions in net profit, answer depends on engineering team strength"
Harvey Bullish Among the first companies to gain GPT-4 access; capable of first-year associate work at large law firms; 6-hour task → 6 seconds
OpenAI Bullish (position action not disclosed) Sequoia investment; ChatGPT viewed as "this generation's Netscape moment"
Hugging Face Bullish (position action not disclosed) Sequoia investment; falls under AI/data theme
Confluent Bullish (position action not disclosed) Sequoia investment; part of the modern data stack
dbt Bullish (position action not disclosed) Sequoia investment; part of the modern data stack
Faire Bullish (position action not disclosed) Founder Max Rhodes known for "continuously applying pressure"
Amazon Positive case study Compared to eBay: business model appeared commoditized, but had Jeff Bezos
eBay Negative case study Marketplace model seemed defensible, but lost to Amazon's team DNA
Workday Mentioned (former Sequoia partner Carl Eschenbach is now CEO) No investment judgment involved

Judgments Worth Remembering

1. "The core of sustainable competitive advantage is not network effects or IP, but team DNA." (Pat Grady) — Amazon beat eBay not because of its business model, but because of Jeff Bezos; DoorDash won not because of its delivery model, but because of Tony Xu. Founder culture creates "advantages that compound over time."

2. "We don't invest in today's market leaders, but in companies we believe will become market leaders tomorrow." (Pat Grady) — Google was not the first search engine, and Flextronics was not the first contract manufacturer. In tech, the number one player disproportionately captures the majority of market cap.

3. "Unique and compelling value proposition" breaks down into two financial dimensions: Unique → Gross Margin (price setter), Compelling → Operating Margin (low sales expense). (Pat Grady) — If the product is unique, you should be a price setter; if the product is compelling, customers should come to you naturally.

4. "Bad numbers can hide a good story — Snowflake's revenue lagged because it closed larger deals, and gross margin declined because globalization exceeded expectations." (Pat Grady) — After Sequoia found all metrics "deteriorating" in Snowflake's semi-annual review, it dug deeper and discovered all good news, increasing its investment from $15 million to $200 million. Key takeaway: "Keep asking until the picture is clear."

5. "Even if current AI model capabilities were frozen, merely optimizing (cheaper, faster, easier) could revolutionize nearly every industry." (Pat Grady) — Current LLMs are 4-6 orders of magnitude less efficient than the human brain (in terms of energy/computation), and nature has proven that better architectures exist. In the long run, more energy-efficient, fundamentally different architectures will inevitably emerge.

6. "Sequoia chooses 27 investors + 65 platform operators, rather than 270 investors — because only two things must be done by people: building founder relationships and making decisions." (Pat Grady) — Spreading knowledge across hundreds makes no one special; concentrating it among the few gives everyone a chance to become "the next Doug Leone."

7. "An organization's output is the sum of individual vectors — vectors have magnitude and direction. Reference checks should determine: Was this person the best at every step? Are they exceptional at what they care about?" (Pat Grady) — Borrowing from Elon Musk's framework. Direction matters more than magnitude: perhaps high school GPA was low, but at that time they were already a genius at entrepreneurship or programming.

8. "What separates legendary companies from the rest is 'sustained pressure' — and what drives you to sustain that pressure depends on your core motivation." (Pat Grady) — Michael Moritz's words to Faire founder Max Rhodes were a watershed moment. Motivation can be obsession with customer problems, a love of building, or a sense of responsibility to employees — but "wanting to issue a press release announcing you're a unicorn" is not enough to sustain it.