Meta Platforms reported first-quarter 2026 results on April 29 that should have been cause for celebration. Revenue hit $56.31 billion, up 33% year-over-year and ahead of the $55.45 billion analysts expected. Net income came in at $26.8 billion. Adjusted EPS reached $7.31 versus the $6.79 consensus.

Then Meta raised its 2026 capital expenditure forecast to between $125 billion and $145 billion, up from its prior guidance of $115 billion to $135 billion.

The stock fell roughly 7% after hours.

That gap between a strong earnings beat and a falling stock tells you something important about where we are in the AI investment cycle, and it matters for every business thinking about how hard to push on AI right now.

Why Investors Flinched

The capex raise was the story. Meta attributed the increase to “higher component pricing” and “additional data center costs to support future year capacity.” That last phrase is the key one.

Zuckerberg has been explicit on this: Meta is building infrastructure today to support value it expects to unlock in future years. He is not claiming that spending $145 billion on data centers will pay off in the next quarter. He’s making a multi-year bet on AI becoming the primary driver of business value, across advertising, consumer apps, enterprise products, and eventually physical hardware.

Investors who expected the AI spend cycle to plateau saw it accelerate instead. That’s what spooked the market.

A 33% Revenue Beat Is Not Enough Anymore

It is worth sitting with that for a second. Meta grew revenue by one-third in a single year. Net profit was $26.8 billion in one quarter. And the stock still fell because investors are worried about what happens next.

That is not a sign that AI is failing. It is a sign that the market has moved past “is this company doing well?” to “is this company building the right foundation for the next decade?” And on that question, the jury is still out. Not because AI doesn’t work, but because infrastructure bets of this scale take years to show up in earnings.

The same tension is playing out at Microsoft, Google, Amazon, and every other major technology company. All of them are spending heavily, all of them are growing revenues at accelerating rates, and all of them are also facing investor questions about when the spending stops and the returns compound.

What This Means for Business

If you are a business owner watching this and wondering what it means for your own AI strategy, here is the honest translation:

The largest, most profitable technology companies in the world are spending at a pace that makes investors nervous. Not because they doubt AI will be transformative, but because they cannot yet model precisely when and how those returns land. They are building now because they believe the alternative, waiting for certainty, is more dangerous than the spend.

That logic applies to businesses of every size.

You do not need to spend $145 billion. But the question Zuckerberg is answering right now: do we build AI capability before the ROI is obvious, or after? That is the same question every business owner should be sitting with.

The companies building data infrastructure, process automation, and AI-native workflows today are not ahead of the trend. They are responding to it at roughly the right time. The companies that wait for the ROI to be undeniable will find the gap between them and their competitors has already compounded.

The Difference Between Cost and Investment

One reason Wall Street reacted poorly is that AI infrastructure spend reads as a cost on a spreadsheet. It shows up in capex. It compresses margins. It delays the moment when earnings improve relative to revenue.

But when it works, it becomes the moat. Meta’s own AI infrastructure is what enables its recommendation systems, its ad targeting, its content moderation, and increasingly its Llama models that enterprise customers are now deploying. None of that is “cost” in any meaningful sense. It is the thing that makes the business defensible.

The lesson for business owners is not to match Meta’s spend. It is to stop treating AI implementation as a cost centre and start treating it as an infrastructure decision with a long payback window.

The companies that understood data infrastructure that way five years ago are the ones getting real leverage from AI today. The companies understanding it that way now will be in that position in 2028 and 2029.

Zuckerberg knows that. The market may be taking time to agree. But history suggests he is not wrong.

---

For a deeper walkthrough of tools like this and how they fit together, the free Working With Claude field guide covers the ecosystem end to end. Get the guide.