How a16z Growth Invests

1. Key Themes

The Evolution from Reactive to Proactive AI Interfaces

The future of consumer AI will shift dramatically from today's chat-based interfaces to proactive, multimodal systems with long-term memory. George argues that while ChatGPT has achieved unprecedented scale (reaching billions faster than any technology in history), the chat interface is "way too limiting" for AI's ultimate potential. The monetization opportunity is vast yet largely untapped - with a billion users but fewer than 50 million paying customers.

"I don't think that the future of how we interact with AI is going to be a chat bot. Like I just think that's way too limiting. I think the big shift will be a sort of what is reactive today to something that's proactive in the future." [00:02:08]

George notes that active users spend almost 30 minutes daily in AI products (compared to 50 minutes on Instagram, 70 on TikTok), creating massive consumer surplus available for capture through new monetization models beyond traditional subscriptions. He predicts something like an "affiliate thing that happens" - native to AI rather than copied from previous paradigms, similar to how feed-based advertising emerged as "probably the best advertisement format in history" despite being unpredictable beforehand.

Growth Stage Investing in the Age of Institutionalization

The venture capital industry has matured into a "grown up" asset class managing $5 trillion in private market cap - representing nearly a quarter of the S&P 500 and more than half the Mag 7. George emphasizes that eight of the ten largest companies by market cap are technology companies, with seven being West Coast venture-backed firms. This transformation demands a more sophisticated, institutionalized approach.

"We're a grown up industry now. Like this is no longer some little bespoke asset class... The private markets have become a real asset class... There's five trillion of private market cap that is, you know, up 10X in the last 10 years." [00:28:23]

The competitive landscape has intensified with multi-stage firms on one side and specialized boutiques on the other. Success now requires years of relationship building rather than sensational one-off efforts. George describes winning deals through "helping them as if we were already investors in their company" over two-year periods, providing candidates, customers, and demonstrating deep business understanding. The growth fund's current portfolio averages 112% growth at 21x revenue entry multiples - metrics George would "do in a heartbeat for the rest of my career."

The Technical Terminator Archetype and Founder Assessment

George has developed strong conviction around a specific founder type he calls the "technical terminator" - individuals who start deeply technical but evolve into commercially excellent business leaders. This pattern creates more sustainable competitive advantages than pure business operators, as technical founders are "likely to figure out the next product area because they're technical, because they're in the products."

"The thing that I like about these technical terminators is they start technical and then you never know if these people are going to become commercially minded, excellent, you know, sort of business people. And then over time they learn the business side." [00:00:09]

Key examples include Ali from Databricks (who wasn't initially CEO and became one of seven co-founders), George Kurtz from CrowdStrike, Dave from Roblox (whose quiet demeanor masks ruthless competitiveness), and emerging AI founders like Michael from Cursor and Shiv from Abridge (a practicing cardiologist who now sleeps at the office). The notable counter-example is Travis from Uber, where the market demanded a "pure battle" mentality to fight mayors and competitors, requiring different founder-market fit.

2. Contrarian Perspectives

The Market Systematically Undervalues High Growth

George argues that markets fail to properly value companies growing above 30% because modeling persistent high growth feels "unnatural." Historical analysis proves this dramatically - in 2009, consensus estimates for Apple in 2013 were off by 3x, "and that's the most covered company in the world."

"I think above 30% growth, the market still doesn't fully value the growth rate... It is just so hard for any investor to build a five or 10-year model where high growth persists. It's just not natural." [00:51:50]

The mathematics are striking: a company growing 80%, then 75%, then 65% produces roughly 3x different valuation than one growing 80%, 65%, 50%, 40%, 30%. Yet the natural inclination is always to model deceleration. This creates systematic mispricing that growth investors can exploit. George prefers investing in 112% growers at 21x revenue over 12% growers at 15x EBITDA because "growth just takes care of so much for you."

Most Enterprise AI Value Will Go to End Users, Not Vendors

Contrary to top-down market sizing presentations claiming AI will capture massive portions of white-collar labor costs, George maintains a skeptical stance on ultimate business models, believing "90% of the technological surplus is going to go to the end users."

"I think when you see major technological shifts, it's very tempting to say, oh my gosh, there is so much economic value that all these companies are going to capture top down. The reality of doing it is much harder." [00:08:38]

He uses the steam engine analogy: it wasn't priced based on replacing 50 laborers - competitive forces drove appropriate returns while productivity gains accrued to end users. This pattern will repeat in AI. Only discrete task completion (like customer support) and consumption-based models (like coding) have clear business model evolution. Everything else remains "pretty TBD." Despite this, the next generation of companies can still become larger than predecessors given capability gains - similar to how cloud companies captured only a portion of value but became enormous businesses.

Waymo Required 400 Cars to Dominate San Francisco

The perception versus reality gap in autonomous vehicles reveals how technology can appear ubiquitous with surprisingly small deployments. Waymo overtook Lyft's market share in San Francisco Bay Area (which has ~50,000 Lyft drivers) with approximately 400 cars.

"They have like 400 [cars in San Francisco]... If your cars are driving optimal routes and sort of fully utilized and not running into some of the problems that drivers have, like it's pretty good." [00:15:33]

This has profound implications for capital efficiency and market structure. George's investment journey with Waymo itself exemplifies contrarian patience - initially resistant in 2020 ("I don't like this at all. Like this is crazy. It's going to take 10 years"), he was overruled by Marc and Ben who insisted "this is autonomous driving. Like, are you kidding me? This is the mother of all markets." By 2024, when cars were actually working and consumer preference was undeniable, the firm invested much more heavily.

3. Companies Identified

Databricks

Description: Data analytics and AI platform built on open source foundations
Why mentioned: Exemplar of the "technical terminator" founder archetype; Ali Ghodsi started as one of seven co-founders, wasn't initially CEO, and evolved into an exceptional business leader who "knows more about like sales ops and hiring processes and reporting lines and all these things you have to do as a manager than probably any of our CEOs"
Quote: "Ali is the technical terminator. Like he self evident. It's self evident. It wasn't self evident, you know, all along. He actually wasn't even CEO. You know, he became CEO later." [00:17:53]

Figma

Description: Collaborative design platform that transformed how designers and engineers work together
Why mentioned: Case study in growth investing requiring nuanced market understanding; demonstrated unique product leading to unique distribution
Quote: "The ratio of designers to engineers is basically double for the modern technology companies. So that's a leading indicator. Everyone is going to, you know, that ratio is going to change. There's going to be double the designers in the world. More importantly, the whole engineering to design process is changing." [00:35:12]

Cursor

Description: AI-powered code editor built on VS Code
Why mentioned: Perfect example of "product fucking market fit" with viral growth and immediate enterprise adoption; demonstrates unique product creating unique distribution
Quote: "We get these notes in like in the case of cursor every single time. It's like immediately to pop. Immediately to pop, proof concept, whatever. Immediately to pop. And like, oh, immediately to full sale deal." [01:02:02]

Abridge

Description: AI medical documentation platform
Why mentioned: Led by Shiv, a practicing cardiologist demonstrating technical terminator qualities, including planning to sleep at the office despite having a family
Quote: "Shiv from Abridge, was practically practicing cardiologist who has then shifted his attention to, you know, building a technology company. He lives in Pittsburgh and he commutes to New York to work most of the time." [00:20:46]

Harvey

Description: AI legal assistant platform
Why mentioned: Demonstrates how reasoning model improvements drove step-change increases in customer engagement
Quote: "Harvey is an example of a company where as the models have gotten better, customer engagement and usage is actually really grown. It actually took kind of a step change... it kind of happened at the same time as reasoning." [00:57:13]

Waymo

Description: Autonomous vehicle subsidiary of Alphabet
Why mentioned: Example of long-term technology development paying off; demonstrates capital efficiency with 400 cars dominating San Francisco
Quote: "Waymo overtook them [Lyft with ~50,000 drivers] in market share... They have like 400 [cars]." [00:15:46]

GitHub

Description: Code hosting and collaboration platform
Why mentioned: Historical example of product so good it sold itself; closed $400k Walmart deal with zero phone conversations
Quote: "We sold to Walmart and they're paying us $400,000 and no one ever talked to them on the phone. We were like, wow, this is an incredibly magical product and an incredibly magical market." [01:00:50]

Roblox

Description: User-generated gaming platform and metaverse
Why mentioned: Led by Dave, another technical terminator who appears quiet but is "ruthlessly competitive" and cares deeply about market cap creation
Quote: "He's the kind of guy that on the surface, if you didn't really know him well, you would kind of be like, oh, is he a little bit, he's a little quieter. And it turns out like he's ruthlessly competitive." [00:20:15]

CrowdStrike

Description: Cybersecurity platform
Why mentioned: George Kurtz exemplifies technical terminator archetype
Quote: "George Kurtz from CrowdStrike is a great example of it." [00:19:49]

Perplexity

Description: AI-powered search and research platform
Why mentioned: George's personal anecdote about using it for deep research on baseball bats revealed the potential for AI-powered shopping and decision-making
Quote: "I did deep research on, you'll probably relate to this, a new baseball bat for my son... If I had to do that on Google, like it would be a total mess... Perplexity was really, really, really good at it." [00:04:46]

4. People Identified

Ali Ghodsi (Databricks)

Description: CEO of Databricks, originally one of seven co-founders
Why mentioned: Perfect embodiment of the technical terminator - started technical, wasn't initially CEO, became exceptional business leader
Quote: "The thing that I like about these technical terminators is they start technical... And then over time they learn the business side. So it's been so fun to work with Ali because he knows more about like sales ops and hiring processes and reporting lines and all these things you have to do as a manager than probably any of our CEOs." [00:18:20]

Dylan Field (Figma)

Description: CEO and co-founder of Figma
Why mentioned: Technical terminator archetype; "one of the nicest guys in our industry. But he is brutally ruthlessly competitive"
Quote: "Dylan from Figma is a great example of this. Like he's so nice. He's one of the nicest guys in our industry. But he is brutally ruthlessly competitive." [00:20:31]

Michael (Cursor)

Description: Founder of Cursor
Why mentioned: Leading "very special founder" aggressively pursuing enterprise while maintaining exceptional product
Quote: "Michael is a very special founder and his team. They recognize what they have and then they are aggressively pursuing the enterprise at the same time." [01:01:22]

Travis Kalanick (Uber)

Description: Former CEO of Uber
Why mentioned: Counter-example to technical terminator thesis; perfect founder-market fit for brutal competitive battle
Quote: "That market was just a pure battle. Like it was like you need to fight. Yeah, like you fight mayors, you fight competitors... you just needed to be ruthlessly competitive and driven and operationally intense. And you know, that's he's the perfect counter sample to that." [00:19:18]

Marc Andreessen and Ben Horowitz

Description: Co-founders of Andreessen Horowitz
Why mentioned: Drive competitive, high-performance culture; made contrarian Waymo investment decision overruling George's initial skepticism
Quote: "Mark and Ben came to me and they said, hey, you know, we got to do this Waymo investment. And I said, I know, like I don't like this at all... They said, you know what? Don't care. Like this is autonomous driving. Like, are you kidding me? This is the mother of all markets." [00:14:16]

Bob Swan

Description: Long-time mentor, operating partner at a16z, former Intel CFO
Why mentioned: Provided valuable advice on calendar management and delegation
Quote: "Bob Swan, who is a long time mentored and friend of mine and an operating partner at our firm gave me this really good advice that he and John Donoho at the end of every year always went through an exercise where they spent like two hours looking at their calendar from the year. And then they had an objective of cutting 30% of stuff." [00:37:44]

5. Operating Insights

The Scoreboard Mentality and High Performance Culture

George instituted specific cultural principles for the growth fund including "there's a scoreboard in this business and our expectation is that we win" and "we are the Yankees and we're going to act like it." The Yankees principle isn't about arrogance but about recognizing the privilege and responsibility of being part of a top-tier brand.

"What I mean by that is we're lucky enough to be a part of a firm that has an incredible brand. And so we're going to run our team very, very high performance. Like if you're on the Yankees, you better be performing. Like this is the big stage." [00:43:45]

New employees must sign both their offer letter and the firm's culture document. For junior team members, "contribution to collective investment judgment" is explicitly part of promotion criteria from day one - unusual for junior investors but crucial for collaborative decision-making. This creates expectations for performance and participation regardless of seniority level.

Single Trigger Puller Decision Process Over Investment Committees

Rather than traditional growth equity investment committees where partners present and vote, a16z growth fund uses single trigger pullers like their venture process. This encourages intellectual honesty and full exploration of both risks and rewards without political pressure.

"We openly expect disagreement. But once you disagree, you disagree, and then you commit. I think by doing it this way, you encourage people to fully explore the risks of investing and fully explore the rewards. You're never in this temptation to sell or to politic for a vote." [00:46:43]

George's first "investment committee" was breakfast with Marc and another partner before officially joining. The process remains informal but rigorous, allowing rapid iteration without Monday committee schedules. The small team of ~10 investors can move faster precisely because they eliminated committee overhead while maintaining decision quality through deep collaboration.

Time Blocking for Deep Thinking in High-Volume Environment

Despite meeting 30 companies per week as a team (George personally meets ~10), he blocks specific thinking time: two hours every Tuesday, two hours every Thursday, plus two 90-minute afternoon blocks weekly. This protected time often gets consumed by pressing needs but the discipline ensures learning doesn't get crowded out.

"I find that I learn a lot and develop a lot of my own thinking just by having think time. And I'm the kind of person that has 20 things open in the browser and I want to read them all and then I don't get to them. So unless I block off a bunch of time, I actually just don't find that I'm spending the time learning as much as I should." [00:39:40]

He abandoned scheduled one-on-ones in favor of spontaneous team contact ("I talk to my team all the time. I'll call them after hours"). The target time allocation is 80% learning about companies and spending time with entrepreneurs, 20% on founder support and internal management - though this shifts during fundraising periods.

Question-Driven Meeting Structure Over Formal Presentations

In initial company meetings, George keeps introductions "super brief" and asks founders to spend just five minutes on strategy and vision. Then he asks questions for 20 minutes based on prior research (website review, customer conversations).

"I've read your website. I know a little bit about the company. I've talked to some customers maybe, but like I need to hear the like the what is the bigger. Like what do you want to tell? You tell me. And then I just ask questions for 20 minutes." [00:40:43]

The ultimate compliment: "thanks, you've done your research" or "thanks for asking that question. That's pretty smart." This approach extracts more signal than formal presentations while demonstrating preparation and thoughtfulness that differentiates a16z from competitors who rely on standard pitches.

6. Overlooked Insights

The Private Markets Now Constitute a Quarter of the S&P 500

While much attention focuses on individual unicorns or IPO markets, George reveals that private technology market cap has grown to $5 trillion - nearly a quarter of the entire S&P 500 and more than half the Mag 7. This represents a 10x increase over the last decade.

"This is something I'm studying now because, you know, the venture industry, you know, is sort of seen as like this small, non-scalable thing. Turns out there's five trillion of private market cap that is, you know, up 10X in the last 10 years. And it's honestly some of the best companies in the world." [00:28:34]

More striking: the public universe in software/consumer/tech has fewer than five companies growing 30%, while a16z's portfolio averages 112% growth with many large enough to be public. The small cap public universe has shrunk by half over 20 years with markedly lower quality than available private opportunities. This structural shift means private markets deserve treatment as a legitimate asset class rather than a niche alternative, yet this realization "hasn't really fully hit the finance industry."

Feed-Based Advertising Success Was Fundamentally Unpredictable

George casually mentions that "we never would have predicted what a feed-based advertisement is. Like no one would have known what that is, because we didn't even know what the feed-based product was. It turns out, it's probably the best advertisement format in history."

"It's not surprising that it monetized this really high. And people actually really like it. Like I really like Instagram ads." [00:04:26]

This observation has profound implications for AI monetization speculation. Just as no financial model in the early 2000s could have predicted feed-based advertising (because the product format didn't exist), attempting to predict AI monetization formats may be similarly futile. The best business models emerge from new product paradigms, not from applying old frameworks. This suggests humility about predicting how ChatGPT's billion users will eventually be monetized - the answer likely involves formats that don't yet exist and wouldn't occur to us today.