Meta Doesn't Know What Business It's In & The Traffic Data Shows It

Search Engine Journal published an analysis connecting Meta’s Q1 2026 user decline with Theodore Levitt’s classic “Marketing Myopia” framework, arguing the company keeps redefining itself without understanding what its users actually want. The piece pairs Julia Angwin’s NYT guest essay (“Meta Is Dying”) with Similarweb traffic data to make the case that the numbers back up the thesis.

What’s actually new

The core data points: Meta’s daily active users dropped from 3.58 billion (Q4 2025) to 3.56 billion (Q1 2026). Revenue hit $56.3 billion, up 33% YoY, but total costs rose 35% to $33.44 billion. Ad impressions grew 19% while ad prices rose 12%, yielding a 27% jump in revenue per user — the classic “squeeze harder” pattern. Meanwhile, the Similarweb data for March 2026 shows Facebook at 11.9 billion monthly visits and Instagram at 7.1 billion, dwarfed by Google’s 86.9 billion. In the AI category, ChatGPT sits at 5.7 billion visits, Claude.ai at 613.7 million (up 423.7% YoY), while Meta.ai doesn’t crack the top 100. The article also notes the counterargument: the QoQ user dip may partially reflect Iran internet disruptions and Russia’s WhatsApp ban, and 4% YoY user growth isn’t exactly a death spiral.

What it means for your config

This isn’t a tooling release or API change, so there’s nothing to migrate or reconfigure. But if you maintain ad-tech integrations, analytics pipelines, or conversion tracking configs that depend on Meta’s platforms, the underlying signal quality question is worth monitoring. The article’s point about Advantage+ performance ($4.52 ROAS per Meta’s own reporting) depending on a healthy, engaged user base is operationally relevant: if user engagement degrades, the behavioral signals feeding your attribution and targeting configs degrade too. There’s no action item today, but teams running heavy Meta ad spend should be watching engagement metrics in their own dashboards rather than relying solely on Meta’s aggregated numbers. The announcement doesn’t detail any platform or API changes — this is purely an analytical piece, not a product update.

Recommended next step

Read the original for the full Similarweb data and the Levitt framework applied to each of Meta’s six pivots. If you’re a team that allocates significant budget to Meta’s ad platform, the practical takeaway is to build your own engagement quality checks rather than trusting platform-reported ROAS at face value. Cross-reference your conversion data against independent analytics. If signal quality is slipping in your verticals, you’ll want to know before Meta’s aggregated numbers tell you.

---

Read the full announcement on Search Engine Journal → Meta Doesn’t Know What Business It’s In & The Traffic Data Shows It