Analysts expect AI to be the central theme yet again during earnings calls for the first three months of this year, with Meta reporting on Wednesday, followed by Microsoft as well as Google owner Alphabet on Thursday. But the talk from executives has shifted from “what can AI do” to “what is AI doing,” and all three of this week’s Big Tech companies will have a different answer. 

For Microsoft, cloud computing platform Azure is the star of its AI portfolio, according to analysts from Goldman Sachs and Citi. Launched more than a decade ago, Azure now offers businesses a suite of AI tools including a service that builds custom generative AI apps from existing models. Azure is “one of the best-positioned AI players” in the market and the “main lever” for Microsoft to make money on generative AI right now, Goldman Sachs analysts wrote in a Sunday note to investors. That confidence partly comes from Microsoft not having to drastically change how it makes money to drive additional revenue from AI, they wrote. 

In contrast, most of Google’s revenue comes from search and related services, which AI from other companies could eventually replace. Google could cannibalize its search business by building a rival AI one, but that would leave it in a precarious position. 

Azure’s business grew 30% year over year for the quarter ending Dec. 31, with six points of that growth coming directly from the demand by its customers for AI-related products, CFO Amy Hood said during the company’s earnings call. Expect an additional one to two percentage points of growth from AI services for the first three months of 2024, Citi analysts said in an April 18 report. Microsoft also has other budding AI businesses that the analysts say will contribute meaningfully to revenue next year—namely the company’s virtual assistant, Copilot. 

Analysts are generally less confident in Google’s business than in Microsoft’s, with Goldman analysts calling its parent’s stock a “wall of worry.” In addition to questioning the growth of Google’s search business in an AI future, investors worry about the company’s operating margins and the impact of potential regulation. Investors will also have “a heightened degree of concern” about how much Google is spending on AI over the coming earnings seasons, Goldman’s analysts wrote. 

Still, Google is a leader in the AI cloud world, which includes many of the services it offers to other businesses, Goldman analysts said in an April 15 report. At Google’s Cloud Next event earlier this month, the company announced expanded access to large language model Gemini and an AI-powered coding assistant. 

While Meta offers an AI assistant and large language model similar to Microsoft’s and Google’s—including its new Llama 3 model introduced last week, which is free for research and commercial use—that’s not what analysts have their eyes on. In notes from Goldman, Citi, and Bank of America, analysts put much more emphasis on how Meta is using AI to improve its social media algorithms. 

Meta, which owns Facebook and Instagram, makes most of its money from advertising. The more time users spend scrolling through their feeds, the more ads they see, and the more revenue Meta generates. In recent months, the company has been working on rebuilding its model for recommending videos to Facebook users, with the help of AI. If effective, the tweak could keep users on Meta’s platforms for longer. When tested on Facebook Reels last year, the new AI-powered technology contributed to “an 8% to 10% gain in Reels watch time,” Facebook head Tom Alison said at a tech conference last month. 

It is still early for these tools, Bank of America analysts said in an April 19 note, but that means investors could enjoy some “positive product surprises” and revenue growth this year. And with a potential U.S. ban on TikTok, a major competitor, Meta’s business is looking well positioned at the moment, they said.