Meta’s central risk is not a lack of growth. It is a measurement gap between the scale of its AI investment and the revenue that investment must eventually produce. The company remains an advertising business: advertising accounted for 97.8% of reported revenue in 2023 3. Family Daily Active People increased 3% year over year in the latest cited data 12. Reality Labs, AI-enabled glasses, and infrastructure partnerships may broaden the platform, but they are not yet equivalent to the company’s advertising engine.
The investment case therefore depends on whether Meta can convert AI spending into measurable gains in ad pricing, conversion, recommendations, and operating leverage. The question is not whether it works, but how you know it works.
Key Insights
Advertising remains the economic foundation
Meta’s advertising franchise continues to provide the cash-generating base for its expansion. User growth remains positive but modest, with Family Daily Active People up 3% year over year 12. Historical ad-delivery growth of 28% shows that Meta has previously increased both engagement and monetization 3. That figure refers to 2023 and should not be projected directly into 2026. Still, the broader advertising market is expected to expand strongly in 2026 8, giving Meta a favorable demand environment.
The market is demanding more than modest user growth can provide. Analysts forecast full-year revenue of $253.71 billion, implying growth of 26.3% 1. That gap means the forecast relies primarily on pricing, ad load, conversion efficiency, AI-enabled targeting, and mix. Meta’s AI program is therefore not simply a technology expense. Its practical purpose is to increase the yield from a very large advertising inventory base.
Relative growth adds pressure. DoorDash’s latest-quarter revenue growth exceeded Meta’s by eight percentage points 6,15,18, while DoorDash reported growth of 36% 5,6,15,18. This comparison does not erase Meta’s scale advantages. It does show that Meta must prove AI is producing durable revenue growth and operating leverage, rather than merely supporting engagement or investor enthusiasm.
AI creates monetization potential and capital risk
The second measurement problem is capital intensity. Management indicated that 2026 capital-expenditure dollar growth would be notably greater than in 2025 7. Meta is participating in a broader hyperscaler infrastructure buildout. Amazon, Alphabet, and Meta issued $169.5 billion of investment-grade bonds in 2026 13, with issuance up 400% year over year 13. A separate estimate places combined debt raised by Oracle, Meta, Alphabet, and Amazon at $194 billion by early July, compared with $108 billion for all of 2025 16.
These figures establish the scale of the commitment. They do not establish the return. AI infrastructure may improve ad ranking, recommendations, creator tools, and generative-product adoption. In the near term, however, the cost appears through capital expenditure, depreciation, energy consumption, and potentially financing expense. The appropriate interpretation is reinvestment for durable monetization—not an assumption that every AI dollar will generate an immediate return.
Investors have so far rewarded the company’s capacity to fund that reinvestment. META shares rose 4.9% amid strong EBITDA and balance-sheet metrics 14, and communication services led the market with a 4.3% gain, supported by Alphabet and Meta 9. Yet average year-to-date earnings growth was reported at negative 15.7% 4. The contrast suggests that investors are looking through current investment pressure toward future AI monetization. That creates undetected risk if revenue, margins, or capital efficiency fall short of the assumptions embedded in the share price.
Reality Labs is an option, not yet an earnings pillar
Reality Labs is showing early evidence that AI-enabled hardware may have greater commercial traction than traditional virtual-reality products alone. Reality Labs revenue grew 16%, driven by increased sales of AI-enabled glasses 17. A separate disclosure reports Reality Labs revenue of $431 million, also up 16% 2. The claims are directionally consistent.
The strategic value of glasses is greater than their current revenue contribution. They could provide a new interface for Meta’s AI assistant, social platforms, and advertising ecosystem. They could also extend consumer-AI distribution beyond the smartphone. The available evidence does not show that Reality Labs is profitable or that glasses growth can offset the cost of Meta’s broader metaverse program. A 16% growth rate is encouraging, but it does not yet make Reality Labs a material earnings engine.
Partnerships add optionality, but not established revenue
Meta’s relationship with AST SpaceMobile is described as an infrastructure-oriented initiative focused on growth and expansion rather than near-term income generation 10. A partnership could strengthen AST SpaceMobile’s strategic position and commercialization prospects 10. Meta may be exploring ways to extend connectivity and platform reach, including in underserved markets.
That is strategic optionality, not base-case earnings. AST SpaceMobile’s strategy requires substantial execution and capital 11. Its risks include supply-chain constraints, financing needs, competition, technology obsolescence, cybersecurity, and uncertainty over whether collaborations become commercial revenue 11. Meta may obtain ecosystem benefits without making satellite connectivity a core operating segment. These relationships should not be capitalized as established revenue streams until binding contracts, deployment milestones, and customer adoption are demonstrated.
Implications for Investors
Meta sits at the intersection of three durable themes: AI-enabled advertising, hyperscaler infrastructure investment, and AI-native consumer hardware. The first is already economically material. Meta’s user scale and advertising reach provide an internal funding mechanism for AI development, while expanding advertising budgets support demand. The second is increasingly balance-sheet intensive. The third may produce a longer-term platform transition, but it remains too small to support the current valuation on its own.
Meta’s strongest path is the least speculative one: embed AI into existing products where incrementality can be tested. Improvements in ad ranking, targeting, recommendation, and conversion offer better near-term visibility than standalone hardware or infrastructure bets. The central risk is attribution collapse between spending and returns—capital expenditure grows faster than monetization, margins compress, and management asks investors to wait for benefits that cannot yet be measured.
That sensitivity is visible in the gap between modest user growth 12 and the 26.3% revenue-growth expectation 1. It is also visible in the reported negative year-to-date earnings-growth figure 4. The positive share-price response to EBITDA and balance-sheet strength 14 confirms that investors value Meta’s ability to invest. It also raises the hurdle for future results.
The relevant scorecard should be concrete:
- Advertising revenue growth and pricing.
- Conversion efficiency and evidence of AI-driven incrementality.
- AI-related return on invested capital.
- Capex as a percentage of revenue.
- Depreciation, cash flow, and operating-margin trends.
- Reality Labs losses and AI-glasses adoption.
- Binding commercial commitments and deployment milestones from infrastructure partnerships.
Bottom Line
Meta’s advertising franchise remains the investable core. AI is the principal mechanism for improving the productivity of that franchise and defending its competitive position. Reality Labs and connectivity partnerships expand the opportunity set, but they remain execution-dependent and should be valued accordingly.
The history of advertising is a history of unmeasured waste. Meta now has the scale and technical resources to measure more of its AI investment than most advertisers could measure in the age of direct mail. The question is whether it will disclose enough incrementality, cost-per-acquisition integrity, and capital efficiency to distinguish productive reinvestment from a larger waste fraction.