The News
Salesforce has introduced outcome-based pricing for its customer service AI, charging companies only when a bot resolves a customer issue without human involvement. The model represents a deliberate shift away from activity-based SaaS pricing, where enterprises pay for seats, API calls, or usage volume regardless of results. Craig Crisler, CEO of SupportNinja, which manages CX operations for major global brands, has publicly questioned the framework, arguing that “the goal isn’t to resolve more tickets; it’s to solve more problems,” and that the real challenge is reaching industry consensus on what a successful AI outcome actually looks like.
Analyst Take
The measurement problem is harder than the pricing model
Outcome-based pricing is a compelling idea on paper. It aligns vendor incentives with customer results, reduces the risk of paying for automation theater, and forces AI suppliers to stand behind their claims. Salesforce is betting that its resolution rate is good enough to make this model profitable. That’s a meaningful signal. But Crisler’s critique cuts to something more fundamental: the enterprise market hasn’t agreed on what “resolved” means, and until it does, outcome-based contracts will be negotiated on contested terrain.
This isn’t a hypothetical concern. Consider the obvious edge case: a customer contacts support with a billing dispute, receives an AI-generated response that technically closes the ticket, but calls back three days later more frustrated than before. Did the AI resolve the issue? Under many contract definitions, yes. Under any honest business measure, no. The pricing model creates a structural incentive to optimize for the metric rather than the outcome, which is precisely the problem Crisler is naming.
Why AI governance is the real battleground
The deeper issue here is one of AI governance, not pricing mechanics. Enterprises rushing to deploy customer-facing AI agents are discovering that defining success criteria is harder than deploying the models. ECI Research’s 2026 DevSecOps and AppSec survey found that AI code governance is the #1 priority investment area for enterprise security teams heading into 2026. That ranking isn’t incidental. It reflects a market that has moved fast on AI deployment and is now catching up on accountability frameworks, measurement standards, and the guardrails that separate a production-grade system from a glorified pilot.
The pilot-to-production gap is directly relevant here. The enterprises that successfully operationalize AI agents tend to be those that define success metrics before procurement, not after. Salesforce’s model rewards vendors that clear that bar, but it doesn’t help customers who haven’t done the definitional work first.
What ITDMs need to understand before signing
For IT decision-makers evaluating outcome-based AI contracts, the key negotiation is not the price per resolution. It’s the definition of “resolution” in the contract itself, the appeals process when outcomes are disputed, and whether human escalation paths are counted as failures or as intended design. A well-designed AI system should route complex cases to humans quickly and accurately. That’s value. A pricing model that penalizes handoffs creates perverse incentives at exactly the wrong moment.
The stakes are higher than they might appear. According to ECI Research’s 2025 AI Builder Summit survey, 44% of enterprise AI leaders have only moderate confidence that AI agents can act autonomously without human intervention. That’s nearly half of the organizations best positioned to evaluate this technology, expressing genuine uncertainty about autonomous action. Salesforce’s pricing model implicitly assumes that autonomous resolution is the desired end state. For a significant share of enterprise buyers, that assumption is still unproven.
Developers building on top of agentic platforms face a related challenge. Instrumentation matters enormously in an outcome-based world. If your integration layer can’t distinguish a genuine resolution from a deflected escalation, your organization is flying blind in a contract negotiation with a vendor whose financial interest is to count every closed ticket as a win.
Looking Ahead
Salesforce’s move will accelerate a broader repricing of the AI services market. Expect competitors and emerging vertical AI players to introduce their own outcome-based tiers over the next two to four quarters. The vendors that win this transition won’t necessarily be those with the highest resolution rates; they’ll be those that invest in shared measurement frameworks, transparent audit trails, and dispute resolution mechanisms that give enterprise buyers genuine confidence in the numbers. The companies that get this right will have a durable pricing advantage. The ones that game the metrics will face contract renegotiations and churn.
For enterprise buyers, the immediate priority is building internal capability to define, instrument, and audit AI outcomes before signing outcome-based contracts. That means cross-functional alignment between CX leadership, product, and engineering on what success looks like, down to the ticket-level definition. The vendors pushing outcome pricing are, intentionally or not, doing enterprises a favor by forcing that conversation earlier. The organizations that treat this as a procurement exercise rather than a governance exercise will find themselves renegotiating from a weak position, while the ones that do the definitional work upfront will hold the leverage.
Stay Ahead of Application Development Trends
Get weekly analyst insights, research notes, event coverage, and AppDevANGLE updates delivered directly to your inbox.
Subscribe for Weekly Insights
Join technology leaders, practitioners, and GTM teams following the trends shaping modern software delivery.
Looking for deeper research access?
Explore ECI Research reports, survey insights, and market analysis through the ECI Research Portal.
