Meta Platforms reports first-quarter results after the closing bell on Wednesday, April 30, and the bar Wall Street has set is high enough that even strong numbers risk a sell-the-news reaction. Consensus is calling for revenue of roughly $48.4 billion, up about 33% year over year, and adjusted earnings per share of $7.65, according to Bloomberg-compiled estimates published Sunday. The shares are up roughly 18% year to date and trade at 28 times forward earnings, leaving little room for a miss on the two metrics that have come to define the AI-era Meta story: Family of Apps revenue acceleration and capital expenditure.
The Llama 5 monetization test
The bull case for the quarter rests on whether generative AI is meaningfully lifting Meta’s core advertising business. The company’s Andromeda recommendation system, retrained on Llama 5, has been the primary commercial AI deployment since its general rollout in February. CFO Susan Li told the JPMorgan Global Technology conference earlier this month that Reels recommendation quality had improved “in the high single digits” and that a portion of that gain was being captured as price-per-impression uplift on Instagram and Facebook (Meta investor relations).
Analysts will be watching three numbers. Family of Apps advertising revenue is expected at roughly $46.8 billion. Average price per ad is expected up 9% year over year, well above the 6% growth rate seen in Q4 2025, with impressions up 5%. And the closely tracked daily active people figure across the family of apps is expected near 3.42 billion, up from 3.35 billion in the prior quarter. A weaker price-per-ad print would feed the argument that the AI ad lift is plateauing; a stronger one would validate the thesis that recommendation systems running on frontier models compound advantage.
The $120 billion capex question
Meta in January raised its 2026 capex range to $115 billion to $125 billion, up from $100 billion previously, citing both training infrastructure and a buildout of AI inference capacity inside the family of apps. That number — implying nearly $30 billion of capex for the March quarter alone — has become the single most contested line in the model. JPMorgan’s Doug Anmuth wrote Friday that consensus assumes roughly 70% of the spend goes to GPUs and AI servers and 30% to data-center shells, networking, and submarine cable. Bears point out that depreciation will catch up: Meta now expects 2026 D&A of roughly $35 billion, up from $24 billion in 2025, a headwind that will compress operating margins by 250 to 300 basis points absent revenue acceleration.
CEO Mark Zuckerberg has been unambiguous that he intends to spend through any near-term margin compression. “If we are going to be wrong about how aggressively to invest, I would rather be wrong by spending too much,” he told employees at an internal AI-strategy review in March, as reported by The Information.
Reality Labs and the Orion question
The smaller but more emotionally charged number is Reality Labs. Consensus expects roughly $1.05 billion in revenue and an operating loss of $4.7 billion, both modestly worse than Q1 2025. Quest 4 launches in the holiday quarter; Orion-class AR glasses remain on a 2027 commercial timeline. The question is whether a path to break-even by 2028 — repeatedly cited by Zuckerberg as the long-run thesis — still survives a year in which the operating loss is on track to exceed $19 billion (Reality Labs segment disclosures).
What would move the stock? An advertising beat with sustained price-per-ad acceleration plus an unchanged capex guide would be the bullish setup. An in-line ad print combined with a capex raise toward $130 billion would be the bear’s quarter. The market gets its answer Wednesday at 4:05 p.m. Eastern.
Discussion
Sign in to join the discussion.
No comments yet. Be the first to share your thoughts.