Meta beats Q1 revenue at $56.3B, but DAP miss and capex hike to $145B weigh

Meta Platforms delivered its fastest revenue growth in five years during the first quarter of 2026, posting $56.3 billion in sales, a 33% year-over-year gain that cleared Wall Street's $55.45 billion consensus by more than $850 million. Adjusted earnings per share came in at $7.31 against a $6.79 estimate, while GAAP EPS reached $10.44 after a one-time $8 billion tax benefit. Yet by the time after-hours trading opened on April 29, Meta shares had fallen roughly 7%, a reaction that captured investors' fundamental unease: every dollar of outperformance was being immediately recycled into an AI spending machine that just got larger. The company lifted its full-year 2026 capital expenditure forecast from $115-$135 billion to $125-$145 billion. It also announced plans to eliminate roughly 8,000 positions, equal to 10% of its 77,986-person headcount, while leaving an additional 6,000 open roles deliberately unfilled. The combination of record revenue, rising total costs, shrinking headcount, and an escalating infrastructure bill is a paradox that Mark Zuckerberg has spent four consecutive quarters building. For now, the ad engine is generating enough cash to make that paradox sustainable, and the balance sheet has the depth to carry it forward.

How Meta's Ad Engine Hit a Five-Year Revenue High While Users Slipped

The operational story of Q1 is straightforward: Meta's advertising engine performed exceptionally well precisely because the AI investments from 2024 and 2025 began compounding in measurable ways. Ad impressions across the Family of Apps grew 19% year-over-year, and average revenue per impression increased 12%. That dual expansion drove Family of Apps revenue to $55.9 billion for the quarter, the real engine of the business, with Reality Labs contributing a relatively modest $402 million, down 2% year-over-year. Meta's total operating income reached $22.9 billion at a 41% operating margin, and free cash flow for the quarter came in at $12.4 billion, a figure that funds a significant portion of the annual capex commitment from internal generation alone.

The number that troubled investors was daily active people: 3.56 billion, down from 3.58 billion in Q4 and below the 3.62 billion that analysts had modeled. Meta attributed the miss to two specific external disruptions: internet outages in Iran stemming from geopolitical events, and a regulatory restriction on WhatsApp usage in Russia. Neither cause is structural, but both underline that Meta's user base, for all its scale, is not immune to country-level shocks. Zuckerberg acknowledged the disruptions without framing them as a warning sign, pointing instead to 4% year-over-year DAP growth as evidence of underlying momentum. CNBC and Yahoo Finance both noted that ad impression growth was strong enough to more than compensate for the user-count shortfall on the revenue line, but that the miss reinforced a latent concern among portfolio managers: Meta's next billion users are in markets that carry political and infrastructure risk that the Family of Apps cannot fully insulate.

The Capex Paradox: Raising the Ceiling While Cutting the Workforce

Meta's decision to raise its 2026 capex guidance for the second consecutive quarter is the central strategic signal of this earnings report. The new range of $125-$145 billion implies year-over-year infrastructure spending growth of roughly 60-70% relative to 2025 levels. Q1 capex came in at $19.8 billion, below the $27.57 billion that analysts had modeled, which is itself evidence that Meta is staging investments deliberately rather than front-loading them into a single period. Total expenses for the full year remain guided to $162-$169 billion, meaning the capex increase is being absorbed within that envelope, not added on top of it.

What is the capex actually buying? Zuckerberg was specific on the earnings call: data center construction sized for frontier AI model training and inference, custom silicon designed to reduce per-token costs on Meta AI, and infrastructure for the Muse family of generative AI models, including Muse Spark. The company ended Q1 with $81.2 billion in cash and marketable securities against $58.7 billion in debt, a balance sheet that makes the spending commitment credible without requiring external financing. Still, raising the capex ceiling to $145 billion raises an uncomfortable arithmetic: Meta is simultaneously cutting 8,000 human workers while committing to build computational infrastructure at a scale that rivals the largest cloud providers. The workforce reduction is not cosmetic. It is, structurally, intended to fund the machines. Operating expenses outside of infrastructure are being compressed precisely so that the infrastructure budget can grow.

Scale AI, Custom Silicon, and Who Gains Market Share

Meta's AI infrastructure buildout is not self-contained. It creates downstream winners and losers across the semiconductor and AI services stack. Alexandr Wang's Scale AI has publicly benefited from large model-training data contracts with Meta, and the company's continued investment in frontier reasoning models extends Scale's runway for high-value annotation and synthetic data work. On the chip side, Meta has accelerated its custom ASIC program to reduce dependence on Nvidia's GPUs for inference workloads, a move that mirrors similar efforts underway at Alphabet and Amazon Web Services. If Meta's custom silicon reaches volume production in 2027 as internally planned, it would materially reduce the per-unit economics that have underpinned Nvidia's gross margins. Broadcom, which provides custom ASIC design services to Alphabet and is in active dialogue with other hyperscalers, is among the clearest beneficiaries of a world in which every major platform company wants training and inference silicon tailored to its specific model architecture and token workloads.

Meta's advertising technology advantage, built on first-party data, closed-loop conversion measurement, and AI-optimized bid management through Advantage+, is more durable than most competitors publicly acknowledge. Snap's 16% workforce reduction and more than $500 million in annualized cost savings announced earlier in April reflect direct pressure that Meta's targeting improvements have applied to smaller platforms. Pinterest and TikTok are both in the process of rebuilding their ad-auction infrastructure to close the performance gap that Meta opened through iterative AI optimization of its delivery systems. The competitive window is not permanent, but for the next 12-18 months Meta's combination of scale data, closed social ecosystems across Instagram, WhatsApp, and Facebook, and AI-optimized auction dynamics gives it a structural pricing advantage that is clearly visible in the 12% price-per-ad growth recorded this quarter.

The Supply Chain Behind a $145 Billion Infrastructure Commitment

A 2026 capex of $145 billion means Meta will be acquiring enormous quantities of compute, networking, power, and physical real estate. The second-order effects run across several interconnected supply chains. TSMC, which fabricates Meta's custom silicon alongside Apple's processors and Nvidia's GPU dies, benefits from the aggregate demand increase, but TSMC's advanced packaging and CoWoS capacity remains the structural bottleneck limiting how quickly any hyperscaler can bring next-generation AI chips to volume. SK Hynix and Samsung, the two dominant suppliers of HBM memory that stacks onto AI accelerators, are already operating near full capacity and have signaled that lead times on HBM3E extend well into 2027. Meta's capex commitment accelerates those purchase orders and competes for the same scarce allocation slots as its hyperscaler peers.

On the power and facilities side, Meta's data center construction program competes directly with Amazon Web Services, Microsoft Azure, and Google for the limited pool of large-scale power contracts and available land near grid interconnection points. Several U.S. utilities have disclosed multi-year capacity commitments to unnamed hyperscalers in regulatory filings; Meta's $145 billion guidance puts it credibly among the largest buyers in that auction. Cooling systems, custom rack hardware, and fiber interconnects create downstream demand for specialized vendors including Vertiv and Amphenol, both of which have cited hyperscaler acceleration as the primary driver of their 2026 forward guidance upgrades. The capex wave is real, broad, and structural.

Zuckerberg's AGI Bet Is Now a Line Item, Not a Talking Point

The Q1 report marks the moment Meta's AI thesis crossed from aspirational narrative into auditable capital commitment. The company's stated goal, building artificial general intelligence and deploying it at scale across its platforms, is now backed by a capital program larger, on an annualized basis, than Google's entire 2023 infrastructure budget. Zuckerberg described the target in concrete terms on the earnings call: Meta AI should be the most-used AI assistant in the world within 24 months. The Muse model family, the Ray-Ban Meta AI glasses currently in consumer distribution, and the Meta Oakley Vanguard hardware platform are all intended as deployment vectors for that assistant, leveraging Meta's 3.56 billion daily users as the activation surface for a product that competes directly with Apple Intelligence, Google Gemini, and Microsoft Copilot.

The risks to this framing are structural. User adoption of AI assistants in social contexts has moved more slowly than the platforms anticipated. Meta AI launched broadly in late 2024 and has not yet demonstrated the retention metrics that would justify a $145 billion infrastructure commitment at full deployment scale. Furthermore, the capex raise is occurring against a macro backdrop in which every major platform company is simultaneously racing toward the same destination, which compresses any first-mover advantage. Meta's distribution moat, 3.56 billion people who open its apps every day across cultures and income brackets, remains its most credible competitive advantage. Whether AI assistant engagement will translate into incremental advertising revenue, new subscription monetization, or a hardware ecosystem revenue stream remains the central open question for investors who are being asked to accept a very large bill on the basis of projected future returns.

Meta's free cash flow of $12.4 billion in a single quarter, sustained within the discipline of flat total expense guidance, means the company can fund this spending level from operations. The after-hours selloff is not a verdict on Meta's future — it is investors re-pricing the timeline on which the capex converts to returns, and signaling that Zuckerberg will need to show AI-driven revenue metrics, not just AI-driven cost reduction, before the market fully credits the bet.