LinkedIn turned half your recruiter into $450M ARR

**The number that matters.** $450M ARR is what LinkedIn just disclosed for two products — one for enterprise recruiters, one for SMB. That's a 0-to-450 line item in roughly 12 months. The agent takes a recruiter's spec, sifts the LinkedIn graph, and surfaces a ranked shortlist with outreach drafts ready to send. The recruiter still pushes the button. Microsoft just productized the half of the job nobody was getting paid full freight for. Inside Microsoft's P&L, this is the proof point that lets Satya keep $190B of 2026 capex on the books with a straight face.

**Why standalone AI recruiting startups just got squeezed.** Mercor (~$500M ARR), Paradox, Eightfold, and the next 40 ATS-adjacent agents raised on a thesis that needed two scarce inputs: proprietary candidate data and recruiter trust. LinkedIn already owns both. Every competitor's sourcing pull degrades a third-party graph; LinkedIn's agent compounds the graph it owns — every accepted InMail, every "open to work" toggle, every endorsement makes the next ranking better. The sourcing tax is zero. The standalones now have a single survival path: go private-network only (alumni, exec, security-cleared, regulated) where LinkedIn's public graph is irrelevant.

**What this does to your hiring spend, today.** Mid-market contingent recruiting agencies charge 18-25% of first-year comp. A $150K hire = $30K fee. LinkedIn Recruiter + the agentic add-on is pacing $11K-$19K per seat per year, unlimited sourcing. One in-house recruiter with the agent now sources what 2-3 used to. If you're a 20-200 person company paying agency fees, your Q3 budget is wrong. If you're a recruiting agency, your client just got pitched a 70%-cheaper alternative by Satya himself on the earnings call — and they were listening.

**The contrarian read nobody is publishing.** This isn't about hiring. LinkedIn is the test case for every Microsoft-owned graph. Outlook (calendar/email), Teams (collaboration), GitHub (code), Dynamics (CRM) are next — Microsoft just proved that graph + agent + an existing buyer = a billion-dollar line item without inventing a new product. Every "AI for X" startup competing with a Microsoft graph just had its TAM cut in half. Q1's $190B capex isn't building models. It's building the moat under products you already pay for. The agent layer is the new compounding margin.

**The window is short and asymmetric.** LinkedIn priced the SMB version low to land seats fast — today it's a margin loss-leader. Within two quarters they'll add seat-tiered "agent autonomy" pricing — auto-outreach, auto-screen, auto-schedule — and rates climb 2-3x. Operators who lock in 2026 multi-year Recruiter contracts at current rates pin their cost-of-hire below market for the next 18 months. Everyone else pays the upgrade tax. The same playbook OpenAI ran on Plus pricing last summer.