From seats to value with agentic AI
AI agents are reshaping how B2B software captures value, pushing vendors to move beyond seats toward usage-, agent-, and outcome-based pricing. The shift demands smarter packaging, transparent value measurement, and hybrid transitions to protect revenue while customers scale automation. Vendors that build telemetry, outcome baselines, and go-to-market muscle now will outpace rivals as agentic adoption compresses seat licenses.
Points clés
- Outcome-based pricing faces immediate friction: 47% of buyers struggle to define measurable outcomes, 36% worry about cost predictability, 25% struggle with value attribution, and 24% cite external factors affecting results.
- Seat-based subscriptions are under pressure, with 40% of buyers citing seat reduction as a primary cost-saving lever, a trend accelerated by agentic AI.
- Packaging is shifting as 48% of IT buyers plan to increase AI/GenAI spending in the next year and 68% of vendors either charge separately for agentic features or bundle them into premium tiers.
- Customers place higher willingness to pay on AI that delivers domain expertise and completes end-to-end jobs, and lower interest in commoditized single-step functions; “seed” limited access in core tiers to drive adoption before monetization.
- Cost-plus is risky for volatile model costs; vendors are countering margin pressure by raising core prices, gating capabilities in premium tiers, adding usage caps, or positioning AI as high-value differentiators.
- Five agentic pricing models are emerging, each with tradeoffs: usage-based (resources), agent-based, usage-based (interactions), outcome-based (jobs completed), and outcome-based (financial results); 91% of IT buyers prefer partially autonomous agents, limiting full human substitution.
- Early mover lessons: a sales SDR provider pivoted from per lead/email/seat to pay-per-qualified opportunity; a CRM vendor is moving from flat per interaction to charging for discrete AI actions; a CX vendor charges only when an AI agent resolves a query without a human, with explicit definitions and arbitration to avoid disputes.
- Hybrid transitions are advised: run new agentic models alongside seats, backed by telemetry and billing to monetize features and provide transparent usage data, then taper seats as agentic adoption grows.
- Monitor leading indicators to pace the shift: rising agentic usage, declining seat purchases, and customer workforce reductions signal healthy revenue diversification and guide de-risking.
- Outlook: agent-based plus outcome-based (jobs completed) will dominate for software vendors, financial outcome pricing will rise among managed-service providers, and pure resource-based pricing will fade as customers demand predictability and value alignment; the work is led by Boston Consulting Group authors including Naila Dharani, John Pineda, Angel Andeyro, Nate Hamming, Yerim Kim, Nipun Misra, and Akash Bhatia.
À retenir
Start simple: package AI that finishes real jobs, “seed” a taste in your base tier, and charge where value is obvious. Pair agent-based pricing with pay-for-completed-jobs, back it with telemetry, and keep seats around just long enough to land the plane without spilling the coffee. If you can’t clearly define “done,” don’t price on outcomes—unless you enjoy arbitration more than revenue.
Sources
Quiz sur le document: 10 questions
