Meta Platforms: the good, the bad and the big AI question

Key takeaways

• Strong core ads fund today, AI needs to fund tomorrow.

• Scam and deep-fake risk plus VR/AR losses weigh on trust, sentiment, and cash.

• Heavy capex now must show up in ad yield, paid features, or higher conversion.

Meta is signalling both strength and strain. Revenue is rising, growth remains, but a one-time tax charge and ballooning spending are raising questions. For investors this means the opportunity is there—but so is the risk. If Meta gets its ambition right in AI and monetisation, it could pay off. If not, the gap between promise and profit could widen.

Growth on paper, pain in the profit line

In the quarter ended 30 September 2025, Meta reported revenue of USD 51.24 billion, up 26% year-on-year. The reported net income was USD 2.71 billion, dropping sharply from USD 15.68 billion a year earlier. The drop reflects a one-time tax charge of about USD 15.93 billion tied to US tax legislation (One Big Beautiful Bill) that hit the quarter’s numbers hard.

On the positive side, Meta’s core “Family of Apps” division (including Facebook, Instagram, WhatsApp) posted growth. But its “Reality Labs” unit (hardware, virtual/augmented reality) still incurred large losses, in the region of USD 4.43 billion for the quarter.

Meta is forecasting fourth-quarter revenue of USD 56 billion to USD 59 billion. It also lifted 2025 capital expenditure guidance to USD 70–72 billion, largely to fund AI compute, servers and data centres that support its ads and product roadmap.

Portfolio angle: signal over noise

Meta’s dominance in social media advertising gives it a strong base. That means the company is not starting from zero. But the burden of future investments, particularly in artificial intelligence (AI) and hardware, raises questions about timing, return on investment and the risk of capital being tied up for long periods.

The question is payback speed on AI and hardware. Timing and return on capital matter more than headline growth. Treat the next 12 to 24 months as an execution window. If AI lifts ad yield and unlocks paid features, operating leverage improves. If monetisation lags, capex crowds out buybacks and keeps margins tight.

Ads still pay the bills

Meta’s ad engine keeps humming. Users grow, engagement holds, and monetisation improves. Average revenue per user (ARPU) rises and daily active people trend higher, pointing to healthier unit economics. The mechanics are simple. Impressions multiplied by price per ad equals revenue. More relevant ads, better targeting, and higher ad load can lift both sides of that equation.

Watch pricing first. It is the cleanest read on value delivered to advertisers. Faster creative tools and automation help small businesses spend more with less friction. If ARPU keeps rising while user growth slows, the pricing mix is doing the heavy lifting. That is usually a positive signal for margins and cash flow, even as the company funds its longer-term bets.

AI spend today, monetisation tomorrow

Meta is spending heavily on compute, data centres, and AI talent. The goal is clear. Better ranking, better ads, smarter assistants, and new formats across Facebook, Instagram, and WhatsApp. In the short run, this is a cost centre. Depreciation and operating costs rise before new revenue shows up.

The near-term test is whether AI lifts ad yield and time spent enough to offset the bill. Medium term, two levers matter. First, paid AI features in messaging and business tools. Second, higher conversion for advertisers that justify price per ad. If those levers move, the spend becomes a return. If not, capex can crowd out buybacks and keep operating leverage subdued.

Trust and Reality Labs: risks that travel

Scam ads and deep-fake videos create reputational and regulatory risk. If users lose trust or regulators tighten rules, the ad model gets harder, not easier. Early signs would be stricter ad reviews, slower approval times, and rising compliance costs.

In parallel, Reality Labs remains a drag. Hardware and mixed-reality push the frontier, but losses are large and payback uncertain. The strategic choice is pacing. Scale only when the use cases are sticky and carry a clear path to gross margin. Until then, investors should treat Reality Labs as an option on future platforms, not a near-term profit source. If management reins in spend or stages milestones, that would de-risk the story without killing the upside.

When growth meets gravity: three risks to monitor

• Monetisation lags. If AI features don’t translate into higher ad yield or new revenue streams, Meta’s spending may not deliver. Warning sign: ARPU or ad unit price growth slows.

• Regulatory or reputational costs bite. Deep-fake scam ads could trigger stricter regulation or advertiser pull-back. Warning sign: increased regulatory fines, decreased ad demand.

• Hardware distraction. Reality Labs losses widen and distract resources from the core business. Warning sign: losses accelerate or strategic pivot without clear path.

Investor Playbook

• Monitor Meta’s next earnings and look for adjusted EPS (excluding one-time tax) as a measure of underlying business.

• Track ARPU growth in the Family of Apps segment, it signals how well Meta monetises users.

• Watch for capex guidance changes: if Meta scales back investments, it may signal prioritisation of profitability.

• Keep an eye on ad spend trends (both global and by region) and any changes in trust/regulation environment around ad content.

Two engines, one crossroads

Meta is clearly operating in two modes: a robust cash generator today and a high-stakes investor in the technology of tomorrow. Its Q3 report showed revenue strength but masked by a massive tax hit and runaway spending.

For investors the questions now are: can Meta convert its AI and hardware bets into meaningful profit? And will its lead in advertising hold as risks mount? The answers matter. The next few quarters will tell whether Meta’s bold ambitions become payoff or burden.