Meta just shipped a free creator AI that overlaps part of our pitch. Here's the honest picture — and how we win the ground Meta can't take.
| Capability | Meta | Us | Verdict |
|---|---|---|---|
| Predict a post BEFORE publishing | None | Core | ✓ Win |
| Real-time virality / crisis alerts | None | Core | ✓ Win |
| Cross-page intelligence + benchmarking | None (own-page only) | The moat | ✓ Strongest |
| Explain analytics in plain language | Free, native | Yes | ✗ Lose |
| Recommend what / when to post | Free, native | Yes | ✗ Lose |
The pattern: basic analytics is now free. Our value is prediction, alerts, and cross-page intelligence — the things Meta won't build.
Matchmaking is the thing Meta is least likely to copy (they won't show one page another's data) and it's naturally viral. So make it the marketing hook and growth engine. But collaboration is occasional-use — it won't carry a subscription on its own. Build daily retention on prediction + alerts; lead the pitch with collaboration; get off Facebook-only data either way.
Here's where we stand and what I think we do, covering everything:
What Meta has actually built: four things this spring. Creator Assistant (June 4) — a free in-Facebook AI that explains a creator's own analytics and tells them what/when to post. Business Agent (June 3) — a customer-service/sales chat bot for businesses (the prompt Lewis saw), already used by 1M+ businesses. Ads MCP (Apr 29) — lets AI assistants like Claude run paid ad campaigns. And Muse Spark (Apr 8) — their own frontier AI model (their answer to GPT and Claude), built after they abandoned the open-source Llama approach; it scored 4th on the main industry benchmark behind Gemini, GPT and Claude, and powers Meta AI for 3 billion+ users. Of the four, only Creator Assistant overlaps us — and only on the basic-analytics part.
How much they're spending, and the deal news: the numbers are staggering. Meta spent ~$72B on AI infrastructure in 2025 and has guided $125–145B for 2026 — roughly double, and more than our entire industry's funding combined. They bought into Scale AI for $14.3B (June 2025), which brought in Alexandr Wang to run their new Superintelligence Labs. And on June 5 the Financial Times reported (this is what you flagged) that Meta is weighing raising "tens of billions of dollars" in fresh equity to fund even more AI — though Meta called that report "pure speculation" and hasn't hired banks yet, so treat it as possible, not confirmed. The honest read: their billions are aimed at OpenAI and Google, at owning the AI model layer — not at building a tool that competes with us. We will never win on money or compute, so we don't try to.
Their future plans — and whether they'll restrict our data: two directions, and the second is what should worry us. First, Meta is clearly going to keep making their own apps smarter and keep creators inside Facebook (they paid creators ~$3B in 2025) — they've said Business Agent will later add "competitive intelligence," and Creator tools will keep improving. Second, and more important for us: they are steadily restricting the organic data third parties like us can access. This isn't speculation — it's on a published schedule. They already removed impressions & page_fans metrics (Nov 2025), are replacing reach/viewer metrics (June 2026), and are retiring whole API versions (v19 in May 2026, v20 in Sept 2026). And they've killed analytics tools outright before — CrowdTangle (their own tool) in 2024, the Groups API in 2024. The pattern is unmistakable: fewer endpoints, stricter permissions, tighter limits, every year. They wouldn't even need to target us — one routine policy change could break our product in a quarter. That dependency is our single biggest risk, and we can't out-execute it; we can only diversify away from it.
The situation in one line: Meta isn't coming to crush us in a fight — but they've made part of our product free (Creator Assistant), and we sit entirely on data they control and are steadily locking down.
What we still own — and Meta won't build: (1) prediction before posting — Creator Assistant only explains the past; (2) cross-page / network intelligence — Meta only ever shows you your own page, never the neighbourhood; (3) real-time crisis & virality alerts; and (4) agencies, who manage many pages, pay the most, and are ignored by Meta's single-page consumer tools. That's a real, defensible business — just narrower than "an AI assistant for your page."
So the plan is two moves, not more features: first, get off Facebook-only data — add Instagram (nearly free, same API), then a genuinely independent source like TikTok or LinkedIn, so no single Meta decision can end us. Second, re-aim the product at the four things above and sell it to agencies, with collaboration/matchmaking as the marketing hook that makes people invite others (the one thing Meta structurally won't copy). We keep the same team and the same tech — we just point them at defensible ground.
And the safety net: even worst case, we've built a reusable machine — team + pipeline + trainable models. If Facebook ever closes entirely, we re-aim it (multi-platform, agency tool, brand monitoring, a data/API business, a new industry vertical, or paid services as a cash bridge). The company surviving is not the question — only which product we point it at.
The one real decision for us: whether our runway and the Nov 28 timeline let us start the second-platform work now, or whether we narrow scope, move the date, or raise to fund it. That's the call you and I need to make together — everything else here is just clearing the path to it.