Google Deep Dive: Will It Shit the Bed?

MARKETS
Google Deep Dive

Today, I’ll prove to you that nobody, and I mean nobody, is better positioned than Google to win the AI race. But if history has taught us something about this company, is that it can’t be trusted.

From being extremely politically biased to censoring accusations of pedophilia to protect certain YouTube stars, Google has a long history of shitting the bed.

So, the question is, will Google repeat the embarrassment of the last few years or emerge as the absolute winner?

All in all, this is a 15-minute, heavily researched, up-to-date snapshot of Google’s present and future, providing all the information you need to decide whether this is a winning business… or a disaster waiting to happen.

But why should you care? Understanding Google’s future is mandatory if you’re an investor. However, Google’s cultural impact makes it mandatory for anyone willing to understand how AI and the world work, as:

• these people hold the key to what content you see when you search the web,

• they also decide what content your kids see in their free time,

• what restaurants appear on your GPS screen,

• and even hold the key to the future of entertainment.

This is the story of the company whose biggest enemy, as you’ll see today, lies within.

The Current Picture

Before focusing on Google, we need some situational awareness regarding markets, which are going crazy.

What is going on?

The Great Rotation

The Russell 2000, the US small-cap index, has grown almost 11% over the last month, while the S&P500, the large-cap index, has fallen 0.43%.

It doesn’t seem that obvious of a rotation, but things become clearer when you look at the Magnificent Seven (Apple, Amazon, Google, Microsoft, Meta, Tesla, and Nvidia), as these have fallen 30% over the same period.

Adding insult to injury, famous investor Tom Lee thinks the S&P500 will be flat for August and that small caps could rise an extra 15%.

Other famous investors such as Steve Eisman, one of the few people who predicted the housing market crash 16 years ago and made huge amounts of money (Steve Carell even portrayed him in the cult film The Big Short), claimed we were in a ‘psychological rotation.’

Money is flowing out of stocks like Google and into small companies. There are many reasons why this may be happening:

• As Eisman mentioned, people might feel uncomfortable with having such a large chunk of their portfolio on 7 companies that, combined, account for 31.3% of the S&P500 index at the time of writing, and with Apple alone being as big as all 2000 small-cap companies in the Russell index combined.

• All the Mag7 companies boast large CAPEX expenditures in Generative AI, with no apparent returns.

• According to prediction markets, the strong inflation data from June 2024 might suggest the start of a cycle of interest rate cuts by the Federal Reserve starting as soon as September, meaning companies more prone to debt, like small caps, will have access to cheaper financing.

• A potential Trump win is seen as bullish for small caps and bearish for large caps. He’s considered a protectionist, which favors companies doing business mainly in the US, even imposing tariffs on international companies that may force countries to retaliate using the big tech companies, like China with Apple, as targets.

Now that the macro is clear, what has Google’s performance been lately?

Not Great, Not Terrible

On Tuesday, Alphabet (Google’s parent company) presented its results for the previous quarter.

In a nutshell:

• Earned $84 billion in revenue, 14% more than last year’s same quarter

• An overall Net Income (profits) of $23.6 billion

• 87% of revenues came from Google Services, with Google’s ad business representing 76% of total revenues

• Google Cloud surpassed, for the first time, $10 billion in revenues and $1 billion in operating income (profits from that particular business)

• YouTube ad revenues continued to increase, reaching almost $9 billion.

• Free cash flow, a key metric that outlines a company's efficiency in generating cash for discretionary spending (and, thus, its robustness to remain in business), totaled $13 billion, adding towards Google’s total cash amount of a staggering $100.7 billion.

Although these values may seem huge, two points on this:

• According to Wiz board member Jason Lemkin, Google just failed to acquire Wiz for $23 billion due to the latter’s fears that the deal wouldn’t get past Lina Khan’s tight grip on big tech M&A.

• Accounts payable (short-term debt) accounted for $6 billion, meaning much of that free cash flow will go to creditors.

But here’s where smiles stop:

• Google’s network business, which earns money from ads appearing on third-party websites, declined for the eighth straight quarter.

• YouTube's revenues increased despite not meeting analyst expectations. However, we must acknowledge that the reported number does not include YouTube subscriptions, as these are diluted into the Google subscription revenue line ($9.3 billion), so investors may be exaggerating a bit.

• Finally, although expected, Google is still unprofitable in several business lines, such as the aforementioned Generative AI business (more on that in a second) and Waymo’s autonomous driving service, without disclosing tangible revenues from either.

However, the biggest concern from investors came from the billion-dollar bet Google is placing on Generative AI.

Google presented another huge CAPEX cost line of $13.19 billion, half of the operating cash flow (the cash they generated from their business operations), most of which was on Generative AI, as Ruth Porat, Alphabet’s CFO, acknowledged: “The primary driver of this, as you know well, is to support the opportunities we see in AI across the company.”

Questioned on this topic, Google’s CEO commented:

Concerningly, this expenditure is only expected to get bigger, as Ruth continued: "quarterly capex throughout the year to be roughly at or above the Q1 level."

And if we take a longer time horizon and Waldomar from KKR’s take on the matter, we should only expect expenditures to explode across all big tech.

• Between 1990 and 2006, the US added 2 GigaWatt (GW) for data centers.

• For the next fifteen years, that value increased to 8GW due to Web 2.0

• Now, their analysts expect 25-30GW of data center power in the next five years, with the International Energy Agency expecting an increase of 540 TWh of yearly energy consumption by 2026 globally, requiring 61 GW of data center power more from 2022 numbers and being much more than what Germany consumes in a year.

Thus, what do we need to take away from all this?

Simply put, they are all in on AI. In other words, their future as a company is completely intertwined with the success—or failure—of their AI efforts and AI as a technology.

Now, considering their massive bet, the question is: How well positioned are they, and what huge risks will they have to manage?

For starters, they have the most challenging roadmap ahead out of all Big Tech.

AI, AI, and AI

Out of all incumbents they are the company most exposed to AI disruptions. The reason is that their biggest cash flow generator, search, is in the direct trajectory of AI progress.

In other words, while Apple or Microsoft’s main revenue lines will most likely continue to be very profitable no matter how much AI develops, Google must unequivocally adapt its $240 billion-plus ad business to the new reality of AI.

In particular, I’m referring to a future where humans do the asking, and AI does the searching, with examples like Google’s AI Overviews, new competitors like Perplexity, or the very recently announced SearchGPT by OpenAI that will make the fight much harder.

If that wasn’t enough, in the process, they need to figure out how on Earth they will monetize a system in which humans no longer see search ads as the AI does the search for them.

While it must be said that these products don’t seem to be a threat today, as Google’s share of the search market continues to grow, this is more due to these products not being good enough yet, pulling users back to traditional search, where Google’s dominance is absolute.

But the promise of a product that searches for you automatically is too strong to ignore, and thus, Google must guarantee that it will offer such a service in the future. And, importantly, that they will continue to be the best.

But is this dependence on AI impacting the other revenue lines?

AI is still Essential.

As we covered, Google also makes money selling smartphones, subscriptions, and cloud services. That said, AI will still play a huge role in these cases.

• Regarding YouTube, Google is rolling out several new AI features to improve user experience but could disincentivize upcoming creators (more on that below).

• As for its Pixel smartphone line-up, AI is considered a key catalyst for terminal upgrading, as this alone is the reason why Apple has grown its value so much since Apple Intelligence’s announcement, as iPhone sales have not been great over the last few years.

• Regarding cloud services, Generative AI is considered a key driver of growth over the next years, as AI data center investment is huge for all cloud providers.

Long story short, not only does Google need to get AI right, but not doing so could be the end, as all its future revenues can be traced to AI one way or another.

Knowing this, what are the reasons to believe or lose faith in Google? All I can tell is that I’m giving you plenty of both.

The Bullish Case

As I said at the beginning, Google has all the cards to win. But as you’ll see in a minute, if there’s one company that can screw things up, it’s them.

The non-AI-only Products

Before moving to the real deal, Google and its AI future, we need to analyze the outlook of its products other than search.

Without a doubt, one of the reasons Google is so praised is how dominant YouTube is.

• YouTube not only dwarfs other streaming services in terms of eyeballs looking at its product (they have 25% of market share),

• but it’s also the most-seen service on TVs.

In other words, regarding screen time, YouTube dominates regardless of the screen type.

Another big piece of Google’s revenues is Google Maps, which, in some accounts, could be at least a $10 billion business considering its high market share (estimates between 50-70%) and researchers estimating it a $20 billion market on ads appearing on user maps.

However, Apple just announced Apple Maps for browsers, which could cause them to lose market share. As for edge devices, Google’s position in smartphones is minor. Nevertheless, it has failed to reach double digits in any meaningful national market.

That said, with AI possibly becoming a prominent catalyst for the adoption of newer smartphone generations, both Google and Apple have a decent advantage over minor players in their path to capture more users by offering better edge AI models (models running inside your phone).

But enough is enough, as it’s about time I tell you why Google’s AI strategy is so powerful.

The AI Cards

AI has three essential pillars: data, compute, and Machine Learning algorithms. Data and compute are essential to train the models, and the choice of algorithm defines how well your model learns.

And from a strategic standpoint, even OpenAI & Microsoft can only dream of being in Google’s position in all three cases.

Regarding data, they have indexed the entire public Internet through Google search (although efforts like CommonCrawl minimize that competitive advantage). But their biggest victory is that AI should slowly transcend into video data over time to the detriment of text.

But why is this, and why is Google uniquely positioned?

While text gives only a high-level view of the world through the lens of human writing, video allows machines to capture low-level details from the world that text could never hope for. For instance, is reading about New York the same experience as seeing a video?

In layman’s terms, video provides machines with the sense of sight (how things look) and temporal awareness (how things evolve), providing much more detail and depth to the representations that the model learns (what the model actually perceives and understands of the world).

And in that terrain, nobody comes close to the amount of data Google has. For example, 720,000 hours of video are uploaded to YouTube daily, accounting for up to 4.2 PetaBytes.

I believe it’s only a matter of time before AI transcends to video; it’s not a question of if but when. And when that happens, Google will have an insane advantage.

Moving on to compute, here’s where things become really interesting to the point of borderline insulting.

Become a paying subscriber of Full Premium package to get access to this post and other subscriber-only content.

Already a paying subscriber? Sign In