Google is eating itself. Meta is catching up. Wall Street is selling both

Revenue up 25% to $127.6 billion. Search queries growing at their fastest pace in five years. Management confident in its strategy. Alphabet’s reward: a share-price plunge of more than 13% after its earnings report, wiping out over $500 billion in market value, according to Bloomberg. Google is successfully drawing people into its AI future. Investors are trying to work out how much cash that future will leave for shareholders.

Another unfamiliar sight is taking shape alongside it. In the third quarter of 2026, Meta brought in $63.06 billion in advertising revenue against $63.35 billion for Google Search & Other. Just $290 million separates them. A first-ever tie between these two advertising engines is within reach. Google’s total ad business, including YouTube and its advertising network, remains comfortably ahead at $82 billion. But Meta has almost caught the search business that built Google’s empire—and its advertising revenue is growing 26% year over year, against Search’s 12%.

JPMorgan’s Doug Anmuth asked the awkward question on Alphabet’s October earnings call: why was faster growth in queries producing slower growth in search revenue? Chief business officer Philipp Schindler cited lower monetization in AI Mode as one reason. Google is actively moving users into an interface that currently makes less money from them than conventional search results did. Schindler also pointed to currency movements and industry-specific trends. Even with those qualifications, the admission about AI was enough to make for an uncomfortable conversation with investors.

Google said lower monetization in AI Mode was one reason, alongside currency and category trends.

Bernstein’s Mark Shmulik puts a finer point on the threat: what happens to the queries with the greatest advertising value? An assistant can guide someone through a purchase in a single conversation, absorbing what used to be several separate searches. Google keeps the user, perhaps even gives them useful advice, and loses some opportunities to sell ads. Management’s answer is that AI Mode generates additional queries early in the shopping process, while its advertising formats are still evolving. The race is between a new user habit and Google’s ability to make money from it.

That is the dilemma behind the Financial Times’s October 29 headline: Google is eating itself, but it could still eat everyone else too. The threat to everyone else is real enough. Google Cloud’s quarterly revenue rose 103% to $30.7 billion. Google has its own models, its own TPU chips, cloud infrastructure and enormous distribution. Analysts see an advantage in its ability to bring down the cost of AI infrastructure. Those assets give it ways to earn from computing and models while it works out how search advertising fits into the new interface.

Cloud’s spectacular growth has yet to cover the construction bill. Alphabet reported negative free cash flow of $9.07 billion for the quarter, CNBC reported. It kept its 2026 capital expenditure forecast at $195–205 billion and warned of a substantial increase in 2027. Jefferies analyst Brent Thill noted that Cloud was the only major business to beat growth expectations. This is an expensive transition: the new business is expanding rapidly, the established one is falling short of forecasts, and the infrastructure has to be paid for now.

Meta comes close to being a textbook example of using AI to improve an existing business. In their October earnings analysis, Baird analysts highlighted improvements to its advertising platform and targeting from AI investment. Revenue rose 26% to $64.6 billion, beating expectations. For Meta, AI is making the ad engine more productive. At Google, its rollout is already opening a gap between growth in usage and growth in revenue.

The market rewarded that relatively straightforward story with a roughly 7% drop in Meta shares on October 22, the day after its earnings release. By October 28, the stock was down about 15.6% from its closing price on the month’s first trading day. AI spending worried investors, even as many analysts raised their price targets. Wells Fargo retained its positive rating but trimmed its target, citing uncertainty over a potential capital raise and spending plans for 2027. An advertising business can beat expectations and still see shareholders sell when the next budget looks more ambitious than the next profit forecast.

We have already examined the economics of this buildout: an accelerator sale immediately brings in revenue for the supplier; the buyer still has to earn back the outlay by putting that computing capacity to work. The test has now reached two of the internet’s most important cash generators. At Alphabet, investors must weigh the advertising revenue the AI transition could consume against what Cloud, models and chips can bring in. At Meta, they must judge how much of the improvement in advertising will be swallowed by its expanding AI infrastructure. Double-digit sales growth settles neither question.

Google has a particularly difficult balancing act. Moving slowly risks losing ground; moving quickly is expensive; and the new version of its own product currently monetizes worse than the old one. It also has the resources to take business from its rivals. Meta has nearly drawn level with Google’s search advertising revenue, yet October’s stock market offered it no celebration either. Both companies could emerge as AI winners with their vast audiences intact. Shareholders still have to find out whether owning those winners will be as lucrative as owning the advertising empires they already had.