Meta Platforms Inc - Ordinary Shares - Class A

Strong core business momentum (Meta Family of Apps) - Q3 total revenue: $64.6B (+26% YoY); Family of Apps revenue: $64.2B (+26% YoY). - Ads: $63.1B (+26% YoY) driven by +10% ad impressions and +14% global average price per ad. - Engagement: Instagram and Facebook saw all-time highs in time spent (Reels and Feed strength); Threads time spent per user +20%+ YoY. - Monetization efficiency improvements from AI-led recommendation and ad ranking systems (e.g., generative retrieval and predictive models).

Key AI/recommendation initiatives supporting growth - Instagram/Facebook gains attributed to upgraded ranking/retrieval architectures and LLM-based content understanding. - Examples cited: - Instagram Reels: relevance launches in Q3 increased Reels time spent by >4%. - Facebook Feed/Reels: generative retrieval architecture rolled out broadly; increased Reels watch time by ~1% and increased “freshness” (nearly 45% of recommended Reels <1 day old, up 11pp from Q2). - Ads: Meta Generative Recommender and ranking improvements drove conversion lift (e.g., ~2% increase conversions in targeted segments on Facebook Feed). - Surprising/positive: AI content understanding models moved onto first-party “Muse model family,” with offline evaluations showing >80% reduction in per-unit processing costs.

Muse (personal agent) shows early product-market fit, but growth is constrained by compute - Muse adoption: >10M people in US & Canada have set up Muse personal agents in ~1.5 months; Muse ranks #1 in App Store “almost every day.” - Usage/retention: ~25% of users use it on any given day; longer use by early adopters shows improved retention (noted as >2x Threads’ initial level / comparable AI apps). - Commerce engagement: >25% of weekly-actives use Muse for shopping; average purchase size >$200; integrations with Plaid/Stripe indicate volumes exceeded expectations. - Big constraint / challenge: Muse is compute intensive and not yet globally deployed; management expects compute efficiency improvements of >10x over the next year to enable scale and profitability as user/revenue lines grow.