Salesforce takes a run at outcome-based Help Agent pricing

Salesforce launched Agentforce Help Agent, a pre-packaged service agent that will be priced at $2 per resolution. The groundswell behind outcome based pricing is accelerating.

Enterprise software vendors like Salesforce are rolling out outcome based pricing to counter worries about token costs. In theory, pricing on outcomes is straightforward, but there will likely be haggling over definitions and fuzzy math. Oracle recently said it has outcome-based models in industries like healthcare and hospitality and is expanding. Hubspot and Pegasystems recently pitched outcome-based models.

Salesforce said its pay-per-resolution pricing is based on lessons from customer implementations. When Salesforce launched Agentforce it tried to price based on conversation, but then simplified the model before landing on a licensing and consumption hybrid approach. Under the latest plan, Agentforce Help Agent customers would only pay when an agent resolves an issue autonomously from start to finish.

The idea for Salesforce is to show ROI via labor costs comparing digital and human labor and gain traction for more use cases. If a customer gives negative feedback for the Help Agent or asks for human escalation, there is no charge. The digital agent will pass customer context to the service team. In addition, both Data 360 and Agentforce are unmetered during the interaction.

• Don't let your vendor's AI strategy become yours
• How to avoid vendor lock-in in the AI age
• BT150 zeitgeist: In search of AI agent ROI
• How Moody's is thinking about data moats, AI strategy, token costs
• CIOs see AI agents as journey, token budgets a concern, says Box CEO Levie
• AI inference costs are going to be a big concern: What's the fix?
• AI Forum 2026: There isn’t an easy button for AI
• Constellation Research's Futures Forum: What CEOs are thinking
• Where’s tokenomics for the rest of us?

Those safeguards for defining a resolution may win over ever-skeptical CIOs. Help Agent is designed for omni-channel deployments and can answer questions and manage cases out of the box. Help Agent landed with a new Agentforce Customer Service portal. Separately, Salesforce launched 12 new preconfigured Agentforce Commerce agents.

As noted previously, outcome based models are going to face challenges.

• What's the outcome? Both parties need to agree on the outcome and metrics.
• How do you personalize that outcome for each enterprise?
• Who audits the value?
• Who bears the risk of costs to deliver the outcome?

Here's a look at Liz Miller’s interview with Pegasystems on outcomes and AI economics.

More Salesforce:

• If CRM Is Dead, How Is the Relationship Still Breathing?
• Salesforce acquires Fin for $3.6 billion
• Salesforce Acquires Fin and Leaves Us Reading Between the Agentic Lines
• Salesforce acquires Contentful in headless content management play
• Salesforce Doubles Down on Headless with Contentful Acquisition
• Salesforce delivers strong Q1, ups outlook