Audit the Capital Expenditure Math
The single most important number in Meta’s report might not be revenue or profit, but capital expenditure (capex)—the money spent on long-term assets like data centers and AI chips. For 2026, Meta has guided for a jaw-dropping $125 to $145 billion in capex,
nearly double its 2025 spending. Wall Street’s patience is wearing thin; investors have shifted from cheering on AI spending to demanding a return on it. A strong revenue report can easily be overshadowed if the company raises its spending forecast yet again. Before reacting to a revenue beat, check the capex guidance. If that number climbs without a clear explanation of how it will generate new income, the market is likely to punish the stock, regardless of how well the core business is doing. The real story is whether the spending is controlled or spiraling.
Verify the Ad Engine's Fuel
Right now, the main proof that Meta’s AI investment is paying off lies within its advertising business. The AI models are what allow for better ad targeting and more effective campaigns, which in turn allows Meta to charge more. To see if this engine is still running hot, ignore the grand pronouncements and look at two specific metrics: ad impressions and average price per ad. In the first quarter of 2026, impressions grew by 19% while the price per ad climbed 12%. This combination is what fuels revenue growth. If either of these numbers starts to slow down, it’s a red flag. It would signal that the AI enhancements are either reaching a saturation point or that the core business funding the entire AI operation is losing momentum. These two numbers are the tangible result of AI, while everything else is still mostly a promise.
Hunt for New, Real Revenue
The long-term bull case for Meta rests on its ability to create new revenue streams directly from its AI investments. This is where you need to listen for specific clues beyond the ad business. First, is there any update on Meta selling cloud computing capacity? Rumors of a potential deal with AI firm Anthropic and the idea of Meta becoming a cloud provider like Amazon's AWS would be a game-changer. Second, look for any data on the monetization of its new AI models, like the recently released Muse Spark 1.1, which is the company's first to be monetized through an API. Finally, check the progress of paid business messaging on WhatsApp. This segment has been quietly growing and represents a concrete, non-advertising revenue stream. Vague statements about the potential of AI are cheap; hard numbers on these new ventures are what signal a real future.
Check the Profitability Gap
One of the clearest signs of the cost of Meta's AI ambition is the growing gap between its revenue and its profit. For the second quarter of 2026, analysts expect revenue to grow by around 27% year-over-year, yet earnings per share (EPS) are projected to be almost flat. This disconnect tells you that the massive expenses are eating up nearly all the gains from the booming ad business. Another metric to watch is the operating margin, which has already started to slip from its recent peaks. This is the financial manifestation of the AI spending spree. While a company can fund growth with lower profits for a time, a sustained trend of flat earnings despite soaring revenues will eventually make even the most patient investors nervous.
Look Past Raw User Numbers
Meta has reported staggering user numbers for its AI assistant, reaching 1.2 billion monthly active users in early 2026. While impressive, this headline figure can be misleading. A user who tries the AI once and leaves is counted the same as someone who uses it daily. The critical audit here is to look for any color or data on engagement and monetization. Is the AI driving purchases? Are users interacting with it in ways that deepen their connection to Meta's apps? Is the company seeing higher retention on Facebook or Instagram among AI users? Right now, Meta AI still lags behind competitors like ChatGPT in dedicated use. High user numbers generate great headlines, but only sustained, meaningful engagement will translate into long-term business value. Without that, it's just another feature, not a revolution.
