The Shift
HubSpot Breeze moved to outcome pricing April 14, 2026 — $0.50/resolved conversation, $1/qualified lead. First major CRM vendor to commit broadly. Others will follow.
The HubSpot move follows Salesforce’s Agentforce per-conversation pricing ($2/conversation, lowered from earlier announcements) and Decagon’s per-resolution model. The structural reason: customers refused to pay seat-based prices for AI that replaces seats. Token-based pricing (OpenAI, Anthropic) leaves the buyer holding consumption risk; outcome pricing flips it. Watch ServiceNow Now Assist and Microsoft Copilot for similar moves through 2026 — both have signaled outcome-pricing pilots.
Why It Signals Maturity
Vendors willing to price on outcomes have confidence their agents work. Token-based pricing disconnects spend from value. Outcome-based is honest; it puts the vendor on the hook.
A vendor offering per-resolution pricing has measured deflection rates across enough customers to predict their unit economics. They’ve solved the eval problem internally well enough to defend the “resolved” definition under contract review. Vendors not yet offering outcome pricing typically lack that confidence — either the agent quality varies too much across deployments, or the resolution measurement is too contested. Use outcome-pricing willingness as a market signal of vendor agent maturity.
Procurement Implications
Outcome pricing requires outcome verification. What counts as “resolved”? Who arbitrates disputes? Early contracts will codify these. Enterprise procurement teams develop new playbooks.
Six contract terms procurement teams need to lock down. Resolution definition (no human contact within 7/14/30 days, no repeat ticket on same intent, customer thumbs-up). Audit rights (vendor’s resolution data, your right to sample and contest). Dispute process (which party arbitrates, what evidence weighs, what monetary recourse exists). Floor and ceiling pricing (commitment minimums plus overage protection). Out-of-scope categories (fraud, abuse, exceeded volumes). Data ownership (your resolution data stays yours and exits with you on contract end). Without these, the first invoice will produce a fight.
Deployment Implications
Outcome pricing aligns vendor and customer incentives. Vendor invests to make the agent actually work in your context. Traditional software pricing doesn’t reward vendor success at your deployment.
Vendor implementation teams now have a P&L incentive to drive your deflection rate up because their revenue depends on it. Expect more proactive engagement: weekly resolution-rate reviews, prompt-tuning sessions, knowledge-base curation help. The flip side: vendors will push to scope intents narrowly to maximize their billable resolutions, and will resist taking on hard intents where deflection is uncertain. Negotiate scope expansions explicitly to avoid being stuck with the easy intents while complex ones get punted to expensive human queues.
Cost Considerations
Run sensitivity analysis on assumed volume. At $1/resolved-lead with 30% qualification rate on 100K monthly leads, you’re at $30K/month — comparable to a seat-based plan for a mid-market team. At higher volumes, outcome pricing can run 2-3x seat-based costs unless you negotiate volume tiers. Conversely, low-volume customers usually save 40-60%. Model both ends before signing.
What to Do This Week
Pull your current AI vendor invoice, calculate effective cost per outcome (resolution, lead, summary), and compare against any vendor offering outcome pricing for the same workflow.
