Wrap, Then Replace: Ema's 'AI Employees' GTM for Eating SaaS and IT-Services Budgets

Wrap, Then Replace: Ema's 'AI Employees' GTM for Eating SaaS and IT-Services Budgets

A new displacement template is crystallizing at the enterprise application layer: wrap the incumbent's installed software first, then replace it — and bill on outcomes, not seats. The freshest proof point is Ema, which raised a $77M Series B led by Creaegis (with Accel, Section 32, and Prosus increasing stakes) as reported by TechCrunch on September 23, 2026, bringing total funding to $140M and more than quadrupling its valuation from its 2024 round.

The motion. Ema deploys "AI employees" — coordinated teams of agents that run multi-step processes across HR, IT, and finance on top of a company's existing applications. Founder Surojit Chatterjee (ex-Google/Coinbase) describes a deliberate two-phase wedge: Ema first "wraps" around the enterprise's existing SaaS, then customers shrink their dependence on those products. His displacement quote is the sharpest articulation of the endgame yet: "Many of our customers are already on the way to replace [large SaaS applications] completely, removing dependency on them, because they are mostly becoming like a database.¹"

The traction that validates it:

• 50+ active enterprise deals; 1M+ active enterprise users; customers include NTT DATA, Hitachi, ADP, PwC, Google, KPMG, Wipro, and Microsoft
• Revenue grown 50-fold over two years; bookings surpassing $150M (multiyear contract value, not ARR)
• Net dollar retention ~180%; more than 90% of customers expand beyond their initial use case
• Gross margins "close to 80%" — notable because it requires less human support as its AI learns from deployments

The pricing experiment. Ema explicitly rejects both incumbent pricing models: "Ema also does not charge customers based on software seats or the number of AI tokens they consume. Instead, Chatterjee said, its pricing is tied to the completion of tasks and business outcomes." That's pure outcome-based pricing sustaining ~80% gross margins — the strongest counter-evidence yet to the margin objection that has pushed the market toward hybrid consumption models (see The AI Agent GTM Shift: The Backlash Against Outcome-Based Pricing and the Rise of Consumption and Hybrid Models, AI-Native Startups Are Abandoning Seat-Based Pricing for Usage- and Outcome-Based Models).

Two more playbook signals in the story:

1. Services firms as channel, not competition. "A lot of the services companies are working with us... They are also dramatically changing or disrupting their own business models because they understand the human-forward model may not be the best model going forward." Ema is eating IT-services/consulting budgets too — the second revenue pool beyond software seats (cf. The Rise of the Forward Deployed Engineer (FDE) as an AI GTM Weapon for the FDE variant of this).
2. Product-first, GTM-second sequencing. Chatterjee says most of the new capital goes to "expanding its go-to-market operations, particularly sales and marketing, after spending its first years largely building the product" — a deliberate order of operations: prove outcome economics and NDR first, then buy distribution. Expansion targets: APAC, South America, Middle East.

Chatterjee also claims the frontier labs aren't direct competitors — Ema orchestrates 150+ models (frontier and open) and competes on domain knowledge, integrations, and orchestration: "Progress in frontier models is actually very beneficial to us." Combined with the same-week model price cuts (Inference-First GTM: Re-Framing Compute as Customer Acquisition Cost (CAC)), the orchestrator's economics keep improving.