#27: Long Google

Two weeks ago, I put 10% of my net worth into Google stock. This is a first for me: while I have held positions in other big tech companies over time, I’ve always shied away from Google because I don’t really understand advertising.¹

In recent months, many other people have also shied away from Google: ChatGPT is eating into Google Search, and Google’s public response has been tepid. Is this a textbook example of the Innovator’s Dilemma? Will Google’s empire crumble?

Such fear and doubt is reflected in the stock: Google is now trading at a 19x P/E ratio, when its historical average over the past decade is 28x, and today’s S&P average is 26x. In other words, the street ascribes a much lower value to Google’s profits than to those of other companies, implicitly anticipating a collapse in Google’s profitability.

But this is myopic, a view far too fixated on legacy conceptions of Google’s Search and advertising business. While the near term is anyone’s guess, the street substantially undervalues the totality of what Google has built, and how that positions Google for the future. My view is this:

1. AI poses threats to Google’s Search business, but they are overrated and solvable;

2. In fact, AI may supercharge Google’s existing Search business;

3. Google is best-positioned to win the AI race;

4. If Google wins the AI race, it may become a $20T+ company in 5-10 years;

5. Oh, and, by the way, Waymo is a trillion-dollar company hiding in plain sight.²

Points three and four in bold are the ones that really matter. The AI community’s AGI timeline is now only seven years out. Most people do not understand:

• These years will pass quickly;

• As we get closer to AGI, trillions of dollars in potential revenue become unlocked. The first firm(s) to the finish line will win the largest economic prize in history.

• Google is in the lead to win.

It’s easy to miss the value. Many investors are bearish on Google because they are fixated on Search as an immutable one-trick-pony, and Search appears paralyzed in a changing world. But Google’s position for AGI is wildly underrated, and it presents opportunities that make questions like whether Search makes money or not unimportant. There is a much larger game in play now. My bet is that Google slowly but surely turns the ship, and in this essay I’ll chart their path from here to a $20T+ world.

Part 1: Search

Many commentators view AI as disruptive to Google Search: people are going to ChatGPT rather than to Google Search because it provides better answers,³ and the answers are exhaustive such that no monetizable click on an advertisement can occur. But this misses a few things:

• Net search volume is still growing. Google’s Search volume increased by 20% from 2023 to 2024. This may feel like a mature industry, but in some respects it is still early! People are still coming online. Software continues eating the world.

• If Search becomes more like a ChatGPT-style experience, that may decrease link clicks, but not necessarily ad clicks: only ~20% of searches show an ad, and fewer yet result in an ad click. Today, most searches are not monetizable at all.

• ChatGPT-style queries and answers may turn out more monetizable than traditional searches because the questions are higher-intent, and the answers surface far fewer links, which better nudges the user toward any displayed link.⁴ The prose of the answer can further nudge the user. As this matures, I’d expect higher click-through-rates/overall value for advertisers.⁵

• Google has the world’s best dataset on queries, ads, and user behavior, and Google’s ads are already partially AI-generated today. The advertiser only has limited ability to provide guidance. Advances in AI further empower Google’s existing advertising flywheel.

• Finally, Google may eventually capture far more value by not getting paid for an ad click, but by closing the loop and offering the product or service that the user is looking for. This enables Google to capture the full amount the user is willing to pay, rather than just the partial margin ceded to an ad click. More on this later.

In short, the future of Search seems to come down to two questions:

1. If ChatGPT offers a superior form factor, can Search move toward that form factor and avoid disruption? I think so, and it seems to already be happening.

2. Can advertising work just as well in that form factor as in traditional Search? Early results suggest yes, and it may work even better. The ChatGPT form factor is more powerful in how it can present the result to persuade user action.

Finally, if there’s a lesson from the last twenty years: whether for countries or big tech companies, betting on the collapse of an incumbent with great momentum rarely works out. Google has colossal momentum — old user habits die hard, and Google’s services are among the most deeply entrenched in the day-to-day lives of consumers.

Part 2: Positioning for the AI Race

But forget about Google’s Search business for a minute, and consider what Google has:

• The most visited website on earth, the default entry-point to the internet for most humans for 25 years and counting;

• The #1 consumer brand in the world;⁶

• Gemini: arguably the best AI models;⁷

• YouTube: the world’s biggest repository of video data;

• Google Search: the world’s biggest store of internet data, having scraped the entire internet for the past 25 years;

• Google Books: the world’s biggest store of published words;⁸

• GMail: the most popular email client with 1.8B active users;⁹

• Google Drive/Docs/Sheets: the most popular workplace suite in the world;¹⁰

• Android: the most widely used mobile phone operating system on earth;

• A mature devices business including phones, laptops, watches, home assistants…

• Google Chrome: the most popular web browser in the world;

• GCP: their own cloud, behind AWS and Azure;

• TPUs: their own chips for machine learning, now used by OpenAI;

• Global data centers representing about $200-290B in investment-to-date¹¹ and another $75B committed;

• $100B on their balance sheet;

• ~$110B in annual operating profit that they could plow into AI if they so wished;

• ~180,000 employees including some of the very best and brightest machine learning researchers and engineers on the planet;

• A truly massive amount of user behavior and ad performance data;

• Endless weird dark horse projects that aren’t even on the public radar right now.

Don’t be distracted by existing revenue or product-in-market. The more you think about Google’s structural advantage in AI, the more staggering it is.¹² They own the whole vertical stack required to win.

The full strength of this competitive advantage against Anthropic, OpenAI and others is yet to become apparent: where other firms top out, Google can keep pushing. Right now, the big AI labs are all focused on making better use of their not-fully-exhausted resources in terms of data, capital, and compute. Therefore, model performance is pretty competitive, and the perceived market leader switches every few months. But eventually, these firms will fully saturate the data, capital, or compute available to them. And however much they may have, Google has a lot more. Similar to how Mistral, Cohere, and others once looked competitive and then couldn’t keep up against superior resources, the same fate may play out at much larger scale — companies worth tens or even hundreds of billions of dollars exhaust their resources while Google’s products and distribution keep improving.

For the last few weeks, Meta has given us a taste of what it means for a trillion-dollar company with conviction to flex its weight: raiding competing labs to the point that OpenAI shut down for a week. Google has barely begun seriously competing; the world will look different when it does.

Part 3: Winning the AI Race

What does it mean to win the race to AGI? There are three important parts to it:

1. Distribution: AGI replaces white-collar labor, which means that many services and products will become commoditized. Commodities differentiate themselves by the quality and efficiency of their distribution (marketing). Without a doubt, Google has the greatest distribution machine in the world.
   
   Today, companies that offer commodity products figure out their gross margins, decide how much money to give up,¹³ give it to Google, and Google provides them some number of clicks, hopefully yielding a positive return to the company. The more lopsided the importance of distribution, the more margin gets reallocated to Google. There’s an argument that marketing is the final industry — human attention the final scarce resource — and Google is not just the leader, but has a powerful structural network effect and moat. That’s a winning position.

2. Verticalization: one of the biggest unsettled questions in AI is to whom the profits will accrue. At every level of the stack, there is margin uncertainty: for example, do profits accrue at the application level? Or will applications be commoditized, and the pricing power will rest with the fundamental model providers? Or with the chipmakers? Who’s going to squeeze whom?
   
   In light of this, I’m inclined to bet on the player that can truly own the full vertical, from chips to end users, rather than on players in never-ending battle with their partners over margin and control. This verticalization doesn’t just provide greater financial safety, but also superior efficiency from coordinated economies of scale than any vertical stacking of patchwork competitor products.

3. End Products: this is the one that people are really missing. In Distribution we spoke about the importance of Google’s distribution machine in a world of commoditized products and services. But if we assume anything close to AGI, then Google shouldn’t collect a fee for making a referral to a third party service provider: AI enables Google to simply provide that service themselves.
   
   For example, if you Google for a hotel today, you encounter a rich cascade of middlemen and economic interactions: you are served an ad to a booking website, which in turn serves ads to numerous hotels, you look through them, you click through screens of primitive upsells for car rentals, trip insurance, etc.
   
   But nobody wants to browse all these websites and rifle through all the options.¹⁴ Google’s AI will obviate this experience. It already knows all your preferences from years of GMail and Search data: it will pick the optimal hotel, discuss with you the services you’re most likely to want, cut out all the middlemen and collect the fees directly. Talk about platform risk — many highly profitable products and services have been built on Google’s distribution platform, but as AI advances, Google will launch their own, superior offerings, and eat those markets overnight. People are already saying that ChatGPT competes with lawyers in the margins: reflect on what AGI really means, then take that to its logical conclusion.

As we get closer to AGI, trillions of dollars of revenue — every white-collar professional service,¹⁵ every software product — will be up for grabs, and Google owns distribution. If you’re a Google exec and you don’t have a vision for Google providing most of the world’s digital services by 2035, then you’re not being ambitious enough.

Importantly, I’m not suggesting that being first to AGI is all that matters. There’s a ton of value unlock along the way. For example, as coding models and computer use models keep improving, they will perform valuable labor at scale, presenting trillion-dollar revenue opportunities even before AGI.¹⁶ If Google has the AI capabilities to seize those opportunities, owns distribution, and owns the full vertical stack such that they have economies of scale and are totally independent of other players, then that seems like a clear winning position. AGI would only make the position stronger yet.

Part 4: The Challenge

Given all this great potential, why is Google not winning more? Why is the adoption of Gemini models so limited? How is ChatGPT taking market share from Google Search? Why are investors not more bullish? From my outsider’s perspective:

• Internal product ownership is far too fragmented. Different teams own poorly divided parts of the same product.

  • Former Googlers have told me that while they have all the resources, the organizational structure makes it far too hard to ship great products.

  • Therefore, large products have incoherent, low-quality user experiences;¹⁷

  • Internal rivalries lead to sub-par, team-over-company outcomes;

• Lawyer-driven-development: excessive fear and caution around launching products due to operating in so many geographies, and AI having unpredictable outputs;

• Managing for near-term shareholder outcomes creates a demand for caution not to jeopardize search;

What’s missing is courage. The past few years have been enormously rewarding to Mark Zuckerberg, Sam Altman, and Elon Musk — these Nietzschean characters with tremendous will to power, who will bet big and hard and take huge risks with asymmetric payoff no matter the scale. These are wartime CEOs, true live players who will reconfigure reality around themselves and who would not hesitate to fight their competitors to death in hand-to-hand combat if necessary.

First and foremost, Sundar must pick up the wartime mantle and act far more aggressively. It is time to truly compete, and the recent quasi-acquisition of Windsurf is a good first step. Beyond that, there are four key steps:

1. Google’s Gemini needs to be front-and-center. It needs to be on the google.com homepage, perhaps where Google Images is now.¹⁸ This needs to launch as soon as possible, no matter what the armies of internal worrywarts say.¹⁹ Google must push, push, push Gemini to supplant ChatGPT.²⁰

2. Google must reorganize to enable itself to ship great AI software quickly. Too many disparate teams are pulled in to work in loose concert, so speed and overall quality suffers. One approach would be to take the best and brightest, fully mirror the OpenAI structure internally, and then aggressively keep growing that team, drawing talent from other divisions as it succeeds.
   
   This is hard! There is lots of corporate inertia against it. And I suspect it is particularly challenging for Sundar because he rose through Google’s complex political culture — and now he must smash and reorganize the structure that once enabled him to succeed. That’s hard on many levels. But it is necessary.

3. The mandate must be explicit and come from the top. The Google bears are fundamentally right that Search over a Big Corpus of Hyperlinks has an expiration date on it. Larry and Sergey were wise to retain super-voting shares — when push comes to shove, they can do whatever they want. They now need to exercise their legal and moral authority as founders to turn the ship.

4. Brace the shareholders. Google’s near-term financial results need to take a backseat to winning the AI race. This is a very simple priority as a matter of expected value. If Search traffic dips, if CapEx gets expensive, if operating profit temporarily shrinks — none of these things matter given the stakes of the game. (And losing the game is a much worse outcome.)

I’m long Google because I believe that these four steps are readily attainable²¹ and will unlock trillions of dollars in value. I don’t know how quickly Google will get there, but its momentum and resources are so great that they should, even if there’s some near-term bleeding. Google’s position is so favorable that they would have to mess it up immensely not to win.

Conclusion

If you believe that AI will be everywhere, then you should bet on the player that already is everywhere. Google is the winner-by-default in this arena,²² and the moment this starts to crystallize, the public markets will react. Surely there may be near-term volatility,²³ and the task ahead of reconfiguring the company around AI requires bold and uncomfortable action, but the long term looks good:

1. Google has everything required to win;

2. They just have to not mess it up;

3. Winning unlocks trillions of dollars in revenue opportunity;

4. Winning provides a moat via a positive feedback loop.²⁴

Right now at $2.1T, none of this is priced in.²⁵ Many investors, when looking at particularly large assets, will implicitly believe that they’re fully valued. It’s hard to believe that a trillion-dollar-thing, with so many smart analysts looking at it, is actually undervalued — it doesn’t pattern-match how we think of a bargain, the small diamond in the rough. And it feels a little crazy to suggest it could be 10x or 20x larger.²⁶ But I’ve seen this play out enough times in my life to know that it happens. We live in an era of returns to scale, and software continues to eat the world. Progress in AI remains rapid, and the economic consequences are, in some respects, simple and undeniable. The prize ahead is the most valuable one there has ever been, and the public has not yet fully internalized this. The race to AGI is the greatest competition of our lifetimes. Good luck to all!

In alphabetical order, thanks to Anon, Archie, Chris, Coby, John, and Paul for discussions over the year-and-a-half that led to this piece.

1. By this I mean that I don’t really have good intuition for the advertising industry as a whole, and while I recognize that Google has set up a valuable and powerful system of network effects, the exact mechanics of them has always been somewhat opaque to me.

2. This article is already thousands of words without touching on Waymo, so and Waymo’s $1T-or-not is marginal to my overall point, so I’ll just put the argument in this footnote:

1. The self-driving tech is here now. This is no longer speculative. It is clear that all driving will become self-driving in the next decade or so. Self-driving cars are magical, and once somebody tries one, there’s no going back.

2. Most of Waymo’s competitors have fallen off, leaving the field wide open for Waymo to lead. Tesla is next-closest, but its self-driving technology still seems less mature than Waymo’s, and its regulatory positioning is also still much earlier.

3. It would take a while for Waymo to roll out a large fleet of their own, but I expect Waymo will partner with automakers like Toyota to give their cars self-driving capabilities. That can roll out rapidly, and capture all the margin as software revenue. Importantly, Tesla would have a much harder time doing this, because it would compromise/conflict with their existing auto manufacturing business.

4. Waymo as the first mover has a reasonable chance of being the market leader, and Uber is currently valued at $200B while having barely even put a dent in the transportation market as a whole. It follows that the bull case for Waymo is a full order of magnitude larger.

3. Any mention of ChatGPT outperforming Google Search deserves a note about Google’s deliberate multi-year kneecapping of search result quality. For the linked article, I think Ed Zitron is otherwise a bad pundit and wrong about lots of things, but this one seemed fine. (Perhaps this is subject to some Gell-Mann Amnesia.)

4. Credit to JerryCap on Twitter, who pointed this out. I liked his note that Google’s TAM is expanding much faster than ChatGPT can take market share.

5. Early press comms from OpenAI and Perplexity, who are looking to monetize via this angle, are also giving this impression. The suggestion is not just that they would compete with Google, but that they can monetize even better. If true, that’s very bullish for Google.

6. Skeptical readers may point out that these “most trusted brand” surveys are going to be fuzzy and unreliable. Fine, but the overarching point still stands: regardless of whether it’s #3 most trusted and #1 most recognized or whatever, they have an extremely powerful consumer brand.

7. At the time of writing, Gemini-Pro-2.5 is #1 for Text, WebDev, Vision, and Search.

8. The most recent datapoint on the size of Google Book is that by 2019, it had apparently scanned in 40 million books in 400 languages. By comparison, the Library of Congress carried around 25 million books at the time.

9. And GMail’s penetration in terms of data is much larger still: keep in mind that even when someone who doesn’t use GMail sends email to someone who does, Google gets a record of that communication.

10. Sources here are a little suspect, but the fact that Google Drive/Docs/Sheets is mostly free and Microsoft Office mostly isn’t makes a huge difference.

11. This gets a little less reliable since this is not clearly broken out in their public filings, but you broadly have the following datapoints:

• $10-15B/year in CapEx for 2006-2019 (~$140-210B total)

• $30B/year in CapEx for 2019-2024 (~$150B total)

• Analysts usually estimate that 70-80% of Google’s CapEx goes into data centers and infrastructure

Taking 70-80% of $290-360B yields a $203-288B range.

12. As a fun thought experiment, consider this: if Thrive and SoftBank value OpenAI at $300B — and that’s while relying on third-party chips and data centers, building out their own brand and distribution, needing to raise many more tens of billions of dollars — by that standard, what’s a fair valuation for Google’s position alone, in expected value terms? A few hundred billion dollars? A trillion? If that seems too high, revisit the list of what Google has, look at OpenAI’s public roadmap, and let it sink in.

13. Of course, this is in zero-sum competition with other companies offering the same competitive product. This means that the amount of money that the company will have to give up rises steadily over time, slowly approaching 100% of the firm’s gross margins. This is the true genius of Google Ads’ model of selling black-box ROAS.

14. So far, I think it’s clear that people prefer the ChatGPT-style experience instead of browsing ugly, clunky websites plastered with ads.

15. The global sum revenue of the major white-collar professional services (legal, consulting, accounting, etc.) is around $7.0 - $7.5T annually.

16. This is partially important to note because definitions of AGI are so fuzzy. I’m pretty sure that ten years ago, people would’ve called what we have today AGI, and I suspect that people will keep finding reasons to not call current-generation AI AGI. The goalposts move as our understanding deepens.

17. Let me vent: just look at this! On Google’s homepage, there’s no way to get to a ChatGPT-style interface. However, in the top-left corner, there’s a little icon for their Search Labs.

What happens if I click on it?

There are pre-seed companies that wouldn’t make this mistake! If you’re going to do user feature-flagging, then hide the button based on feature-flag status, rather than showing the button, letting the user click it, and then get to a disappointing “sorry!” screen. I’m sure that hundreds of thousands of people are hitting this every day. This is some of the most valuable screen real estate on the planet. It is unbelievable that it is wasted like this.

18. Moving Google Images back a little bit in terms of prominence may be a good thing: Google Images has degraded severely in quality over the past few years. As far as I can tell, that’s because of SEO spam polluting the search space. It’s much harder for me to find an appropriate image on Google Images now than it was ten years ago.

19. The “AI Mode” product that is being piloted with certain users is a good first step.

20. Maybe this will look similar to how Microsoft has used its whole Office/Teams suite to squeeze Slack out of its market.

21. I always like investing when the overall sentiment is crowding out the facts: when people are bearish but there is only a very small number of things that need to go right, perhaps take the bet that those things can go right! People like to imagine that mismanagement is boundless, but it’s corrigible. I moved back to San Francisco with a similar thesis: sure, the city has problems, but all of them boil down to one or two simple governance issues, which are totally fixable. This bet has worked out pretty well over the past few months.

22. The other dark horse here is Microsoft, which also has enormous distribution and could stand up world-class AI capabilities overnight. Meta is trying, but their distribution is much less suitable.

23. Some people think that Search volumes may decrease, and the public markets might get scared. Perhaps there’s a 30% stock price dip on the horizon; who knows. This kind of thing is difficult to time. My view is that I’m happy to ride it out, and if the price dips, I’ll probably double down on my position, barring no changes to this thesis. Even aside from AI, I think the durability and stickiness of Google’s offerings is underrated.

24. I didn’t discuss this elsewhere: the positive feedback loop is usage. Every time someone uses Google’s AI products, they should improve by virtue of having more data, and it also reinforces distribution: the user is going to Google, having a fine experience, and will therefore go to Google next time as well. Being the first port of call is very sticky.

25. Google’s market cap is about $2.1T, of which ~$100B is cash-on-hand, and $500B is YouTube. The remaining $1.5T seems to be valued almost entirely on Google’s Search business, which mostly discounts all the other assets. (Remember our thought experiment from footnote 12: what’s the strategic position for AI worth just on its own?) And while it’s hard to value Calico, Wing, and the many other subsidiaries, it’s clear that Waymo’s $45B valuation will increase significantly from here.

Semi-related, Waymo and Search are somewhat marginal to my bull-case scenario, but they do provide some downside protection and contribute in expected-value terms. The “worst-case” scenario that I can envision is that Google doesn’t get there quickly enough, but the momentum of Search would still sustain some revenue growth such that a medium-term collapse in stock price seems unlikely to me. I think this is an asymmetic bet.

26. Forecasting at this level of scale gets a little fuzzy — we’re talking about AGI-empowered Google substantially eating up software and professional services. Nobody knows what revenue multiples or margin compression would like. Further, if we start to seriously automate big chunks of the labor economy, then lots of dynamics will change in weird ways, including e.g. the value of money itself, but that’s neither here nor there for the purposes of this post.