Meta reported ad revenue growth of 24% year-over-year in Q4 FY25, driven by an 18% surge in ad impressions traced to its new Lattice and GEM machine-learning models. Management highlighted that AI-powered redistribution of ad spend has increased average revenue per person (ARPP) across the Family of Apps by 100% over the past five years. In its Q1 FY26 guidance, the company signaled a continuation of this momentum, forecasting mid-20% revenue growth in its core ads business despite investments in AI infrastructure.

Daily active users climbed 7% year-over-year to 3.58 billion in Q4 FY25, underscoring broad adoption across Facebook, Instagram, WhatsApp and Messenger. Instagram Reels watch time in the U.S. jumped 30% over the same period, supported by advanced sequence-learning algorithms that tailor content feeds and boost session lengths. Executives noted that deeper interaction histories and recommendation enhancements have driven a 15% increase in time spent per user since Q4 FY24.

Capital expenditures are slated to rise to between $115 billion and $135 billion in 2026—a roughly 70% increase from 2025 levels—as Meta ramps up data center build-outs and custom AI-training hardware deployments. While this spending will pressure free cash flow—projected at roughly $5 billion for FY26—CFO commentary emphasized that operating margins (excluding the hardware division) are expected to expand by 200–300 basis points over the next two years once new facilities reach full utilization.

The Reality Labs segment recorded a loss of $14.9 billion in FY25, but management believes those losses have peaked. The next wave of product launches—including smart glasses with integrated mixed-reality capability—is expected to drive initial consumer adoption. Analysts project that once annual production reaches 10 million units and software ecosystem revenues begin to accrue, Reality Labs could contribute positive operating income by FY29, unlocking a new long-term growth driver.