Google Didn't Diversify Away From Ads. It Bought Better Places to Sell Them.

In August 2015, Larry Page renamed his company after the alphabet and told the world that Google was now just one letter in something bigger.⁴ The implication landed instantly: self-driving cars, life-extension labs, internet balloons — Google was finally becoming a portfolio, finally outgrowing the search box. Ten years and tens of billions of dollars later, here is the part nobody put in the press release. In 2023, of every dollar Alphabet earned, advertising still supplied roughly 78 cents.⁸ The famous diversification produced one half of one percent.¹

The official story is that Alphabet expanded beyond search to escape its dependence on ads. The real story is the reverse: Google took the cash an ad business throws off and used it to buy the two places people would spend the most attention — phones and video — so that more attention would flow back through the same ad machine. It didn't diversify away from advertising. It built a wider funnel into it.

The pie chart that gives away the whole game

Read Alphabet's 2023 10-K like a confession rather than a brag. Total revenue was $307.4 billion. Google Search & Other alone was $175 billion — about 57% of everything. YouTube ads added another $31.5 billion, a further 10%.¹ Stack those, fold in the rest of the ad surfaces, and you reach that ~78% number that has barely twitched in a decade.⁸ The most striking line, though, is at the bottom of the page: Other Bets — every moonshot, every Waymo and Verily and balloon — generated $1.53 billion combined, and lost $4.095 billion doing it.¹² A company spending billions to look diversified, while a single product line still carries more than half the weight.

So if the moonshots aren't the expansion, what is? Look at the two acquisitions everyone remembers — and notice what they have in common. Google bought Android in 2005, a deal so quiet its price was never officially disclosed; the often-repeated $50 million is an estimate from unnamed sources, not a number Google ever confirmed.⁷ A year later it bought YouTube — and here too the legend is slightly wrong. It was not $1.65 billion in cash. It was $1.65 billion in stock, a paper deal struck in October 2006, with YouTube left to run independently afterward.⁶ Neither was a bet on a new revenue model. Both were bets on a place to put ads.

Why phones and video were never a detour

Here is the mechanism, worked down to the floor. An ad business doesn't really sell ads — it sells attention, and attention has to live somewhere. In 2005, Google's terrifying risk was that someone else would own where attention was about to move. If a rival controlled the phone, it could set the default search engine, the default browser, the default map — and quietly route a billion people away from Google before they ever typed a query. Android was insurance against that future: give the operating system away free, make Google search the default on it, and the threat of being locked out of the next computing platform simply evaporates. The product is free; the distribution is the asset.

YouTube runs the same logic in a different medium. Video was the one form of attention search couldn't capture — you don't Google a song, you watch it. Buy the place people watch, and you've extended the ad machine into the one room it couldn't enter on its own. That's why YouTube ads now throw off more than $31 billion a year, sitting inside the same Services segment that earned nearly $96 billion in operating profit in 2023.¹² Android and YouTube look like diversification. They are the opposite: they are how the ad business defended and widened its own territory. The expansion fed the core instead of replacing it.

But isn't Google Cloud the real escape hatch?

The honest objection is Cloud — and it's the one part of the story that genuinely cuts against the thesis. Google Cloud is not an ad surface; it sells compute and software to enterprises, and it has grown into something real. It climbed from about 4.3% of revenue in 2018 to 10.8% in 2023.⁸ And in 2023 it crossed the line that matters: its first full year of operating profit, $1.716 billion, after a $1.922 billion operating loss the year before.² By 2024, that profit had more than tripled to $6.112 billion, with Q4 revenue up 30% to $12 billion.³ This is a true second engine being built — slowly, expensively, and finally working.

But notice the timeline. Cloud only became profitable in 2023 — nearly two decades after Google entered the ad business, eight years after the Alphabet restructuring that was supposed to herald diversification. And even now, every dollar of Cloud profit is dwarfed by the $121 billion Google Services threw off in 2024.³ Cloud is the exception that proves how singular advertising really is: it took eighteen years and oceans of patient capital to build one adjacent business that doesn't lose money, and it still amounts to a rounding error against the core. Diversification, when it finally arrived, was glacial — not the bold pivot the Alphabet branding promised.

Even the founding myth of Alphabet itself bends under scrutiny. The 2015 restructuring is remembered as the diversification milestone, but Page's own announcement framed it as a management problem — 'more management scale, as we can run things independently that aren't very related' — not a plan to wean the company off ads.⁴ And it wasn't even a clean CEO handoff: Sundar Pichai became CEO of Google in 2015 but only became CEO of Alphabet in December 2019, when Page and Brin finally stepped away, writing that the company 'no longer need[ed] two CEOs and a President.'⁵ The new corporate skin let the founders run experiments at arm's length. It never changed where the money came from.

Alphabet spent a decade and a fortune teaching the world a lesson it didn't mean to teach: that a great advertising business is almost impossible to outgrow, because the thing it sells — human attention — keeps moving, and the cheapest way to keep selling it is to own the next place it moves to. The moonshots were never the expansion. They were the cover story. The real expansion was a phone and a video site that quietly made the ad funnel wider, and it worked so well that ten years on, three out of every four dollars still arrive the exact same way. Google didn't escape its core. It got better at the same thing, in more rooms.