I'll research META before writing the preview. Let me start by gathering recent filings, transcripts, news, and price data.Let me read the most recent earnings release and transcript, and pull the stock price trajectory.Let me read the Q1 2026 earnings call transcript for management's strategic commentary and any forward guidance detail.I have strong material. Let me get consensus expectations and any final pre-earnings analyst positioning via web search.# Meta Platforms (META) — Q2 2026 Earnings Preview
Meta enters this print with a business that is firing on all cylinders but a stock that isn't. The fundamental debate has narrowed to a single question: will the ~$135B of AI capex the company is pouring into infrastructure this year actually generate a return — and can the ad engine keep growing fast enough to carry the depreciation, financing costs and Reality Labs losses that come with it? The setup is unusually tense because Alphabet just raised its 2026 capex guide days ago and the stock reaction was poor, priming the market to punish any Meta capex hike.
| Metric | Consensus | Context |
|---|---|---|
| Revenue | ~$60.2B (Street), guide $58–61B | +~26–27% YoY vs. $47.5B in Q2'25 |
| EPS | ~$7.13–7.23 | Roughly flat vs. $7.14 a year ago — the key tension |
| Advertising revenue | ~$59.0B | +~27% YoY |
| Family of Apps revenue | ~$59.6B | +~26% YoY |
| Reality Labs revenue | ~$0.44B | Still deeply loss-making |
| Implied Q2 capex | ~$34B | More than double the ~$16.5B in Q2'25 |
The estimate range is wide (Street EPS spans roughly $5.76–$8.55), reflecting genuine uncertainty about margin trajectory. Note the revenue-vs-EPS divergence: ~27% top-line growth but ~flat EPS. That gap is the story — it quantifies how much AI/infrastructure buildout, R&D talent costs and RL losses are compressing what were ~43% operating and ~39% net margins a year ago.
Meta has an exceptional beat history — roughly 15 straight revenue beats and 13 straight EPS beats — so a top-line beat within/above the guide range is close to the base case. The reaction will hinge on the outlook, not the quarter.
The price action tells you the market has already partly "sold the capex." That cuts both ways: a clean quarter with disciplined guidance could trigger a relief rally, but a capex raise into an already-jittery tape is the clear downside catalyst.
Management raised FY26 capex to $125–145B at Q1 (from $115–135B), attributing most of the increase to higher memory/component pricing plus some incremental data-center cost. The setup is treacherous: Alphabet's recent capex hike was "panned," and analysts warn that any Meta raise could spark similar selling. Meta faces a higher bar than Microsoft or Amazon because it has no cloud revenue to offset the spend — every incremental dollar must be justified through the core ads business or nascent AI products. Watch for (a) any change to the $125–145B range, (b) the first real color on 2027 capex, which management pointedly declined to frame at Q1 while flagging it "continues to underestimate" its compute needs, and (c) commentary on memory pricing as a cost driver.
A cluster of recent developments reframes Meta's infrastructure strategy: - Selling compute externally: Reports of an ~$10B, two-year compute deal with Anthropic, and the hire of senior AWS executive Dave Brown to run infrastructure — signals Meta is seriously weighing a cloud/"neocloud" business to monetize its buildout. Zuckerberg has said a cloud business is "on the table." - Off-balance-sheet financing: A $14B El Paso data center with BlackRock, reports Meta needs to raise hundreds of billions for the AI buildout, and a $107B step-up in contractual commitments disclosed at Q1. - Rising cost of capital: Bond investors are demanding higher yields on Meta's AI-linked debt, and Moody's warned that "unprecedented" AI spending threatens credit quality at the hyperscalers. Expect questions on financing structure, interest expense, and free-cash-flow trajectory (FCF was only $12.4B in Q1 as capex ramped).
Any concrete framing of a cloud/compute-leasing revenue stream would be a positive because it directly answers the "where's the ROI?" question.
The bull case requires evidence that AI spend is already feeding revenue. Watch for updates on: - Core ad-system gains — at Q1, adaptive ranking (a ~1T-parameter LLM-scale model), GEM architecture advances and the value-optimization suite (>$20B run-rate, doubling YoY) were driving conversion-rate improvements. Continued conversion gains and higher price-per-ad are the cleanest ROI evidence. - Business AIs — >10M weekly business-AI conversations at Q1 (up from 1M to start the year), still largely unmonetized. Any move toward a monetization model matters. - Meta AI / Muse — engagement lift from the Muse Spark model; the product roadmap for personal and business agents. - New revenue levers — Instagram reportedly plans to charge for in-app AI tools (a subscription-style model), a potential new lever beyond advertising.
Family DAP was 3.56B in March, dipping slightly QoQ on Iran internet outages and the Russia WhatsApp restriction — management said underlying growth was positive ex-those items. Watch for a return to sequential DAP growth, continued Reels/video time-spent gains (Q1 saw ~8–10% video-time lifts), and any read on ad pricing/macro. Impression growth (+19% Q1) and price-per-ad (+12% Q1) are the two levers to track.
Meta cut roughly 8,000 roles (~10% of staff) in May and has framed a "leaner operating model" to help offset infrastructure spend. FY26 total-expense guide is $162–169B (unchanged at Q1), and management reiterated 2026 operating income above 2025. Watch whether the expense range holds and how the headcount reductions flow through margins — plus fallout from a lawsuit alleging AI-driven, discriminatory layoffs.
Bull scenario: Revenue beats toward/above the $61B high end, capex range held (not raised), expense guide intact, operating margin holds up better than the flat-EPS consensus implies, and concrete AI/cloud monetization signals. Given the de-rated ~17–18x multiple and 8% implied move, that combination sets up a relief rally.
Bear scenario: Any capex raise (memory-price-driven or otherwise), a soft 2027 capex framing that implies another step-up, margin compression worse than feared, or fresh legal/financing overhang. With Alphabet's capex reaction fresh in mind, the market is primed to sell a spending increase first and ask about ROI later.
Bottom line: The quarter itself will almost certainly look strong on the top line — the ad machine is working and AI is demonstrably improving it. The stock's direction will be decided by capex, the financing/ROI narrative, and 2027 signposts, not by the Q2 revenue or EPS number. This is a guidance-and-narrative event more than a results event.
(This preview is for informational purposes only and is not investment advice.)