Bernstein Analysis: Advertising Revenue Grows 27%, When Will Meta's Personal AI Monetize?

TL;DR
· Meta's Q2 revenue increased to $60.8 billion, with a Bernstein report maintaining an outperform rating but lowering the target price to $800.
· Advantage+ annualized revenue run rate exceeds $75 billion, but the 2026 capital expenditure guidance remains high at $130 billion to $145 billion.
· Advertising and merchant tools have benefited from AI, while the personal AI entry point, monetization speed, and regulatory pressures remain valuation boundaries.

Meta's Q2 earnings report continues to demonstrate a strong advertising base, but AI investments have also pushed profitability pressures to the forefront. According to a Bernstein report, Meta maintains an Outperform rating, with the target price lowered from $850 to $800.

This downgrade is not a bet on weakening advertising business. Meta's total revenue for Q2 was $60.801 billion, a year-on-year increase of 28%, with advertising revenue of $59.363 billion, up 27% year-on-year. The issue is that the company is keeping AI infrastructure investments at a very high level, with 2026 capital expenditure guidance reaching $130 billion to $145 billion, including principal payments on finance leases. Management has also indicated that AI infrastructure investments will continue at least until 2027.

According to the report, the $800 target price corresponds to an approximate 22 times price-to-earnings ratio for 2027. This target price still retains upside potential, but the premise is clear: AI must not only enhance advertising efficiency but also help Meta maintain the next generation of consumer entry points and gradually find ways to charge for consumer-grade AI.

Meta's stock price has fallen from a high of $796 to about $586 over the past year, lagging the S&P 500 by 18.2% YTD.

Advertising is Accelerating, AI Has Entered Revenue

Meta's hardest numbers in Q2 still come from advertising.

The company's total revenue was $60.801 billion, with advertising revenue of $59.363 billion. Family daily active users reached 3.6 billion, Instagram daily active users reached 2 billion, and Threads monthly active users exceeded 500 million. WhatsApp peaked at processing 30 million messages per second, indicating that Meta's traffic base in social, communication, and business messaging scenarios remains solid.

AI is no longer just a long-term product narrative. The Advantage+ advertising suite has an annualized revenue run rate exceeding $75 billion, with over 9 million small businesses using at least one AI advertising creative tool. For Meta, this is more direct than launching a standalone AI application, as AI is being embedded in ad delivery, material generation, conversion optimization, and merchant tools, entering the company's most profitable systems.

After integrating Muse Spark, Meta AI saw a 60% increase in daily interactions. Business Agents have over 1 million merchants using it weekly. FoA Other revenue exceeded $1 billion in Q2, compared to $583 million in the same period last year, a year-on-year increase of about 73%.

These figures cannot yet be directly equated to new profit curves, but they at least indicate that AI functions are being actively used by users and merchants, rather than just remaining at press conferences and roadmaps.

Meta's Q2 advertising revenue grew 27% year-on-year, remaining at a high level among digital advertising peers.

Zuckerberg Bets on Personal AI to Defend the Next Generation of Entry Points

Zuckerberg's proposal for "personal superintelligence" sounds distant, but within Meta's history, the logic is not complicated: what Meta fears most is not the short-term fluctuations in advertising growth, but the rewriting of consumer entry points by new product forms.

TikTok has forced Meta to passively chase in the short video migration. After the explosion of ChatGPT, Meta also failed to secure a leading position in the minds of the first wave of generative AI users. For a company that profits from user engagement and ad distribution, if users in the future spend more time with AI assistants, AI agents, AI hardware, or new content generation tools, Meta cannot just be a bystander.

This explains why Meta is simultaneously advancing multiple lines: Muse Spark 1.1 paid API, enterprise AI agents, Meta One subscription service, AI-integrated smart glasses, and a commercial tool platform for merchants.

However, these still belong to early signals. Other revenue is growing rapidly, but the base remains small. Whether APIs, subscriptions, revenue sharing, and AI agents can form stable income still requires more time to prove.

FoA Other revenue exceeded $1 billion in Q2, a year-on-year increase of about 73%, but still in the early stages of scaling up.

Target Price Lowered, Pressure from Higher Capital Expenditure

The main pressure for the target price downgrade comes from the investment side.

Meta's Q3 revenue guidance is $61 billion to $64 billion, corresponding to a year-on-year growth of 19% to 25%, which includes about a 1 percentage point headwind from foreign exchange. The revenue side remains strong, but expenses and capital expenditures continue to raise market expectations.

The full-year expense guidance is $165 billion to $169 billion. The 2026 capital expenditure guidance is $130 billion to $145 billion, with the lower end raised by $5 billion from previous guidance, including principal payments on finance leases.

According to this report's model, the EPS expectation for 2026 has been lowered to $32.27, and for 2027 to $36.68. The revenue story has not been overturned, but the profit pressure is clearer. Investors need to believe that these investments in AI servers, models, chips, and talent will ultimately translate into higher advertising efficiency, stronger user engagement, or new consumer AI revenue.

This is also the boundary for the $800 target price. The report uses a 6.7 times EV/Sales for 2027 and a DCF model weighted equally, with DCF assumptions including a 10% WACC and a 3.5% perpetual growth rate. Since public institutions' target price metrics are not entirely consistent, $800 is more suitable as a valuation metric for this report rather than a consensus market expectation.

The 2026 capital expenditure guidance is $130 billion to $145 billion, with capital intensity expected to rise above 50% and remain high in 2027.

Advertising Can Prove the Present, Personal AI Has Yet to Prove the Future

The most convincing part of Meta currently is that AI has already improved advertising efficiency and user experience. Advertising revenue, user scale, and AI advertising tool revenue run rates all support the judgment that the fundamentals remain strong.

However, the personal AI entry point has yet to provide equally strong evidence.

The product forms of consumer AI are still changing. It could be a chat assistant, a content generation tool, AI glasses, or an embedded messaging and business messaging agent in Instagram, WhatsApp, and Facebook. Which one will become a high-frequency entry point, which one can charge steadily, and which one can enhance advertising conversion, remains to be determined.

Regulation will also constrain the imagination space. Meta still faces lawsuits and investigations regarding youth protection, antitrust, and more. If AI products further enter communication, content recommendation, advertising delivery, and scenarios involving minors, compliance costs may continue to rise.

Therefore, the $800 target price retains a conditional optimism: the advertising core business can still provide cash flow for AI investments, but personal AI has not yet turned into a new revenue curve that can be directly written into the profit statement. Meta can continue to spend money to buy tickets for the next generation of entry points, while the market will watch when these investments will no longer just be defensive costs.