Meta Platforms enters its next earnings report with investors focused on two central lines in the model: advertising revenue and capital expenditure. Advertising remains the engine of the business, while spending on artificial intelligence infrastructure continues to rise.

The key issue is whether Meta’s recovered ad machine can comfortably fund an AI buildout that has continued to expand in scope, timeline and budget. The setup is straightforward: strong advertising demand is meeting an aggressive investment cycle. The question for shareholders is how long that balance remains acceptable.

Meta’s current strategy closely resembles its historic playbook: improve ad performance, generate surplus cash, and reinvest that cash into the next platform shift. Today, that shift is AI. The investment is not only about research or consumer-facing features. It is also aimed at business drivers such as better targeting, higher relevance and lower unit costs across the company’s systems.

At the same time, investors still remember Meta’s last major spending cycle. Reality Labs losses have not disappeared; they now sit alongside AI infrastructure spending. Because Meta reports most social-app activity inside Family of Apps while separating Reality Labs, the earnings debate often turns on whether the core advertising business can absorb losses and investment elsewhere without weakening consolidated margins. The market is more likely to tolerate large checks for GPUs and data centers if advertising growth remains stable and AI features begin appearing in revenue lines rather than only in product demonstrations.

Where Meta’s ad growth is coming from

Meta’s revenue still comes overwhelmingly from advertising across Facebook and Instagram. The most important developments are less about new formats than about pricing, infrastructure and the systems that determine ad performance. Three areas are especially relevant this quarter.

Reels is moving from time sink to revenue line

Reels previously diluted monetization per minute. Over the past year, Meta has repeatedly said the monetization gap between Reels and Feed or Stories has narrowed as auction models and ad formats improved through training cycles in its advertising systems. Recent earnings-call commentary has suggested that the gap continues to close, though it is not fully closed. Watch-time trends and pricing commentary are therefore likely to be assessed together.

Messaging remains an underappreciated growth lever

Click-to-message ads across WhatsApp, Messenger and Instagram DM have been described by Meta as a multi-billion-dollar run-rate business and have been highlighted repeatedly on earnings calls, including in transcripts published by The Motley Fool. The model is based on measurable conversations that can lead to conversions, particularly in markets where many storefronts are phone-first. Automation is also relevant, including AI chat flows that can handle the first exchanges with a customer before a human takes over.

Signal loss has been offset by AI-driven tools

Meta’s Advantage+ tools have helped rebuild ad performance after Apple’s App Tracking Transparency changes weakened some data signals. Dynamic creative, broad targeting and model-driven budget allocation now carry more of the workload, according to Meta for Business. Advertisers may place less emphasis on explicit targeting controls when outcomes remain consistent or improve. That shift matters because Meta’s ad system depends less on a single deterministic signal when models can infer intent from broader behavioral and creative patterns.

AI spending: from GPUs to Llama in the feed

The capex number attracts attention because it is easy to track. The underlying story is broader. Meta is investing in training clusters, inference capacity close to users and the software stack needed to turn compute into stronger recommendations and better advertising outcomes.

The capex curve and its components

In 2024, Meta raised its full-year capex outlook to prioritize AI infrastructure, pointing to heavier investment in data centers and silicon through Meta Investor Relations. External reporting at the time emphasized a multi-year, GPU-heavy buildout as the company scaled recommendation systems and model training, with Reuters among the outlets covering the broader AI infrastructure push. That stance has not softened. Larger models and growing inference demand continue to increase the cost of the buildout.

Llama is no longer only a research project

Meta’s Llama 3 release showed a commitment to open-style models and developer reach, not only internal tooling, according to the Meta AI blog. The strategy can benefit Meta by allowing a wider ecosystem to test and improve models while the company applies those gains in ranking systems, safety layers and ad products. The operating loop is straightforward: better models can improve content relevance, more relevant content can support engagement, and engagement can support ad performance.

How AI spending affects the financials

Initial capital expenditure goes into buildings, networking and accelerators, appearing first on the cash flow statement. Depreciation then flows through the income statement over a period of years, widening the difference between GAAP earnings and free cash flow.

Operating expenses can also increase through hiring and model-operations work, including evaluation, red-teaming, safety, trust and support. Cost of goods sold can face pressure as inference scales, because personalization and generative AI features are not free to serve. The potential payoff would appear mainly in two places: higher advertising yield and new revenue lines such as enterprise tools, API usage or subscription value-added services.

Margins, buybacks and investor patience

The trade-off is familiar: invest through the cycle or optimize margins in the near term. Meta’s recent approach has been to invest first and use buybacks to support per-share metrics when cash generation is strong. Timing is the central issue. If capex continues to rise while advertising growth settles closer to the mid-teens, the operating margin path may still appear manageable, but investor tolerance can narrow when guidance feels open-ended.

The margin sensitivities under scrutiny

Two sensitivities drive many financial models: ad pricing momentum relative to user-time growth, and the depreciation curve tied to the AI buildout. Stronger Reels monetization and messaging adoption can offset some pressure. However, another capex tranche for new data center footprints would extend the depreciation tail. Guidance that links spending directly to AI-driven revenue opportunities is likely to be interpreted differently from broad language about investment.

Reality Labs remains part of the picture

Even if investors are more forgiving toward AI than toward VR, consolidated results still include all business units. Any widening of Reality Labs losses is likely to draw questions. Meta has described VR and AR as longer-cycle bets, but the tolerance range narrows if AI capex also stretches. Analysts may press management for clearer hurdle rates, milestones or spending discipline.

Items to watch on earnings day

The earnings narrative can change quickly once results and guidance are released. Several items are likely to receive close attention:

• Advertising growth excluding currency effects: A meaningful slowdown would increase scrutiny of other line items.
• Capex guidance versus prior commentary: Any change in range or language can materially affect how investors assess the spending cycle.
• Reels monetization versus time spent: Continued narrowing of the monetization gap would support the advertising story; a stall would raise margin questions.
• Click-to-message updates: Specific adoption examples are more useful than general references to strong interest.
• AI product connections: Examples where AI directly improved an ad KPI or created a revenue opportunity are the most relevant.
• Operating expense discipline: Hiring pace and areas of focus may provide more information than total headcount alone.
• Buybacks and cash flow: Strong free cash flow can make heavier investment easier to absorb.

Why crypto and fintech audiences are watching

Meta’s AI infrastructure spending matters to crypto and fintech for two reasons. First, it draws on the same supply chains that Web3 companies and quantitative trading firms increasingly use for inference and model-assisted trading. Scarcer high-end accelerators and more expensive cloud inference can affect budgets across the sector.

Second, Meta’s push around Llama can help startups and protocols integrate AI into wallets, support flows and creator tools without depending on a single proprietary vendor. That can accelerate practical uses of AI inside Web3 products.

There is also an advertising angle. When brands concentrate spending on performance channels that can prove conversion, gaming companies, NFT projects and fintech apps with clear cost-per-acquisition metrics may compete for incremental budget. If click-to-message commerce becomes more automated, more payment-flow experiments could move closer to embedded finance. In some regions, crypto on-ramps could sit near chat-based shopping flows, even if Meta remains cautious on custody.

Scenario map for the quarter

The quarter does not require precise forecasts to assess. The main paths depend on what management reports about advertising growth, capex and measurable AI benefits.

Risks and what could go wrong

Several risks could complicate Meta’s earnings setup. A macro advertising slowdown could arrive just as AI spending ramps, squeezing margins from both sides. Inference costs could run hotter than expected if new AI features receive heavy usage without offsetting monetization. Regulatory pressure around data usage or competition policy could complicate model training and ad targeting.

Supply-chain or energy constraints could delay data center buildouts and push spending into less efficient alternatives. Reality Labs losses could overshadow the AI narrative if hardware cycles disappoint. Creator-economy fatigue could reduce Reels time spent and weaken pricing power.

The largest risk is not only a single weak quarter. It is a guidance reset suggesting that the capex curve will keep rising while revenue tailwinds fade.

Frequently asked questions

Why are investors so focused on Meta’s capex guidance?

Capex is the clearest proxy for Meta’s AI ambitions and has a long-term effect on margins. Higher capex generally means more depreciation later, and investors want evidence that the spending will translate into better ad yield or new revenue lines rather than only larger computing clusters.

What would show that AI spending is working?

Concrete links would include improvements in ad conversion and pricing attributed to model upgrades, lower content moderation costs from better classifiers, or early revenue from AI tools offered to businesses. Specific and repeated examples carry more weight than broad claims.

How important are Reels this quarter?

Reels is important because it absorbs substantial user time. If monetization per minute continues to catch up with Feed and Stories, the business can grow without requiring a sudden increase in total time spent.

Is messaging commerce a real business line or just hype?

Messaging commerce is meaningful in regions where shoppers prefer chat-based buying. Meta has repeatedly highlighted click-to-message ads as a multi-billion-dollar run-rate category on calls, and AI-assisted chat flows can improve conversion and response times, which advertisers monitor closely.

What could be viewed as an upside surprise?

Flat or lower capex guidance paired with stronger ad pricing would be the clearest favorable surprise. Metrics directly tying AI features to revenue, along with steady buybacks, would add support.

Does Reality Labs still matter for the stock reaction?

Yes. Even if investors are more receptive to AI spending, widening Reality Labs losses can affect valuation if shareholders believe multiple large bets are crowding near-term returns.

Where can official commentary on Llama and Meta’s AI plans be found?

Meta’s AI blog covers Llama releases and research direction at Meta AI, while Meta Investor Relations publishes guidance and earnings-call materials.

Disclaimer: This article is provided for informational purposes only. It is not offered or intended to be used as legal, tax, investment, financial or other advice.