In partnership with
The Three Numbers That Will Decide This Earnings Season
AI Spending Is About to Face Its First Real Earnings Test
Amazon, Microsoft, Alphabet, and Meta are set to spend a combined 725 billion dollars on capital expenditure this year, up 77 percent from last year's already record high. That figure alone tells you something has shifted. Companies do not commit this kind of money without expecting a very specific kind of return, and this earnings season is where that expectation gets tested in public.
For years, investors accepted vague promises about AI transforming the business. That patience is running out. What matters now is whether the spending shows up as actual revenue, actual margin and actual advertising performance. Three numbers will decide how this quarter is read and none of them are complicated to understand once you know where to look.
Number One: Cloud Revenue Growth Versus Capex Growth
In the last earnings cycle, Microsoft's AI segment hit an annualized revenue rate of 37 billion dollars, up 123 percent year over year. That sounds impressive until you set it against Microsoft's own capex, which is now tracking toward 190 billion dollars for the year. When spending grows faster than the revenue it is meant to justify, the gap becomes the story investors focus on, not the growth rate in isolation.
The same test applies to Amazon Web Services and Google Cloud. Alphabet's cloud revenue reached 20 billion dollars last quarter, beating expectations by nearly 2 billion dollars, and that beat is exactly the kind of signal markets want to see repeated. If cloud growth accelerates in step with spending, the market will read the capex as investment. If it lags, the same numbers get read as risk.
Number Two: Advertising Revenue as the Reality Check
Meta's advertising business remains the clearest test of whether AI is improving a product people already pay to use. Meta's revenue grew 33 percent last quarter to 56.3 billion dollars, its fastest growth since 2021 and the company credits AI-driven ad targeting and recommendation systems directly for that gain. This is the one place where AI spending has to show up as measurable dollars, not future potential.
Alphabet faces a similar test through Google Search and YouTube advertising. Global ad spending is projected to cross 1 trillion dollars for the first time this year and any hyperscaler that cannot show its AI tools are capturing a growing share of that pool will struggle to justify its infrastructure bill in the same breath.
Number Three: Margin Compression From Depreciation
Every dollar spent on a data center eventually shows up on the income statement as depreciation. As capex has scaled into the hundreds of billions, the depreciation expense tied to that spending is starting to compress operating margins, even at companies reporting strong revenue growth. This is the quiet number that does not make headlines but shapes every earnings call transcript this quarter.
When Meta raised its 2026 capex guidance last quarter, its stock fell 6 percent in after hours trading, even though revenue beat expectations. Microsoft and Amazon saw similar declines tied to spending increases rather than performance misses. That reaction pattern is likely to repeat this quarter, and it tells you the market has shifted from rewarding ambition to demanding payback timelines.
Six people doing the work. Your headcount is one.
Your finance close runs in #finance. Stripe and QuickBooks reconciled, runway updated, posted Sunday night without you asking.
Engineering review lands in #eng. Viktor pulled the open PRs, left comments on auth-refactor, flagged a dependency blocking api-pagination.
Campaign brief lands in #growth: Meta CPA up 18%, recommendation to pause broad match, a draft landing page already deployed for the variant test.
You hired him on day zero. He lives in Slack and Microsoft Teams alongside your contractors and investors, connects to 3,000+ tools, pushes back when you ship something dumb.
"Viktor is now an integral team member, and after weeks of use we still feel we haven't uncovered the full potential." Patrick, Director, Yarra Web.
What This Quarter Actually Tests
None of these three numbers exist in isolation. Cloud growth without matching capex discipline signals inefficiency. Advertising gains without AI-driven acceleration suggest the investment isn't translating where it counts. And margin compression without offsetting revenue growth eventually forces a change in spending behavior, regardless of how confident a CEO sounds on a call.
This is the first earnings season where these companies cannot simply describe AI as a long-term bet. J.P. Morgan's own analysts now put combined 2026 capex estimates at nearly 700 billion dollars for just five companies, a figure that has grown by 173 billion dollars since the year began. That kind of upward revision only makes sense if the underlying business case keeps proving itself quarter after quarter.
What to Watch on the Calls
- Whether cloud revenue growth rates match or exceed capex growth rates for the same period.
- Whether advertising executives can attribute specific revenue gains directly to AI tools rather than general market growth.
- Whether any company signals a slowdown or plateau in capex guidance, which would be read as either discipline or doubt depending on the framing.
The Real Test Has Arrived
For most of the past two years AI spending got a pass because it was framed as building for the future. That framing has an expiration date and this earnings season is close to it. The companies that can show their spending converting into cloud growth, advertising performance and manageable margins will be rewarded. The ones that cannot will find that investor patience, unlike capital expenditure, does not compound indefinitely.