Salesforce Reshapes AI Pricing Through Agentforce’s Pay-for-Resolution Approach

Salesforce launched Agentforce less than two years ago as a platform designed to help businesses build, deploy, and manage AI agents. Since then, organizations have created service agents capable of handling a wide range of customer interactions. However, building these agents traditionally required significant effort, including connecting individual communication channels, organizing business knowledge, and defining the actions an agent could perform.

Salesforce is now simplifying that process with the introduction of Agentforce. This autonomous AI service agent comes with a guided setup experience and can be deployed across multiple customer touchpoints within minutes. Built on the Agentforce 360 platform, Help Agent is designed to support customers throughout the entire service journey, from the initial question to final resolution.

The Help Agent is built around three key capabilities that are essential for effective customer service. First, it understands the business by using the organization’s existing Salesforce knowledge. This helps reduce one of the biggest challenges faced by AI agents: poor-quality, disconnected, or unorganized data.

Second, it is designed to take action instead of simply providing answers. Businesses can use its built-in library of actions to manage cases, schedule appointments, update orders, and support other common service processes. Additional actions can also be configured based on each organization’s specific requirements.

Finally, the agent can be activated across multiple channels from a single setup experience. Businesses can deploy it across voice, websites, customer portals, and messaging platforms without having to manage each channel separately.

Salesforce has also tested the technology in its own operations. Agentforce has reportedly handled 4.3 million inquiries on help.salesforce.com and successfully resolved 70% of them. The experience and insights gained from these interactions have helped shape the capabilities of the Help Agent.

One of the most significant changes introduced with Agentforce is its pricing model. Businesses pay only when the AI agent successfully resolves a customer issue independently from start to finish. If a customer needs to be transferred to a human agent or leaves dissatisfied, the business is not charged for that interaction.

This means organizations can align their AI investment with actual outcomes instead of simply paying for usage or activity.

A New Approach to AI Pricing: Pay Only for Successful Resolutions

Salesforce’s pay-per-resolution model is designed to connect AI costs directly to customer service results. Businesses are charged only when the Help Agent independently completes a customer resolution.

If a customer requests assistance from a human service representative or does not receive a satisfactory resolution, no resolution fee is charged. In cases where human escalation is required, the service team receives the complete context of the customer’s interaction, helping them continue the conversation without starting over.

The model also eliminates concerns around unexpected usage costs. During an interaction, businesses do not have to monitor separate metering for Data 360 or Agentforce. As a result, companies can focus more on the results delivered by the AI agent rather than forecasting consumption or managing unexpected overage costs.

Faster Deployment Across Digital and Customer Service Channels

Built-In Knowledge and Agent Testing

Poorly organized or fragmented data can significantly impact the performance of AI agents. To help businesses overcome this challenge, the Help Agent is designed to use existing Salesforce knowledge as its foundation.

Businesses can further expand the agent’s knowledge by uploading additional files or providing web URLs for content crawling. The setup experience also includes an agent preview window, allowing users to test the agent before moving forward with deployment.

This gives organizations an opportunity to evaluate how the agent responds, identify potential gaps, and make adjustments before making it available to customers.

Simplified Omnichannel Deployment and Ready-to-Use Actions

Deploying AI across several customer channels can traditionally involve complex integrations and technical coordination. Agentforce Help Agent simplifies this process by allowing businesses to enable voice, web, portal, and messaging channels through a unified setup experience.

The agent also includes a selection of preconfigured actions that allow it to handle common customer service requirements from the start. It can manage cases and respond to customer questions without requiring every action to be built from scratch.

Businesses can also expand their capabilities by adding processes such as appointment scheduling, order management, and account updates. These additional actions can be configured through existing Agentforce Builder tools or with the help of the customer’s preferred coding agent.

A More Interactive Customer Portal Experience

Salesforce has also redesigned the Agentforce Customer Service Portal, one of the primary channels through which the help agent can interact with customers.

The updated portal centers around a single conversation interface. As customers explain what they need, the experience dynamically adapts by presenting relevant answers, information, and interactive cards. These features allow customers to complete tasks directly within the conversation instead of navigating through multiple pages or systems.

Because the experience can use real-time data, it can also support proactive workflows and engage customers before an issue becomes a larger problem.

Making Service Agent Deployment Easier and More Efficient

Salesforce has also recently announced a definitive agreement to acquire Fin, an AI customer service platform focused primarily on small and medium-sized businesses. Fin provides autonomous, end-to-end AI service agents and is reportedly used by more than 30,000 businesses worldwide to manage customer inquiries.

Once the acquisition is completed, Salesforce customers will have additional options for deploying AI agents across their customer service operations.

The combination of Salesforce’s capabilities and Fin’s experience could be particularly valuable for SMBs looking for faster implementation, easier integration with existing systems, and quicker, measurable results from AI-powered customer service.

The transaction remains subject to customary closing conditions, including the receipt of required regulatory approvals. Salesforce expects the acquisition to close during the fourth quarter of its fiscal year 2027.

Salesforce’s Vision for AI-Powered Customer Service

According to Kishan Chetan, Salesforce’s EVP and General Manager of Agentforce Services, the future of AI-powered customer service is about more than simply responding to customers faster.

The broader goal is to resolve customer issues completely across every channel, ideally during the first interaction. Agentforce Help Agent is built on Salesforce’s experience from more than two years of working with AI agents in real-world environments.

The platform combines Salesforce’s unified and secure architecture with a more personalized, proactive, and omnichannel customer experience. Its guided setup also makes the technology easier for businesses to deploy.

With the pay-per-resolution pricing model, Salesforce is also connecting its commercial success more closely to the success of its customers. Businesses pay when the AI agent successfully resolves an issue, creating a model focused on measurable outcomes rather than simple usage.

Conclusion

Salesforce’s Agentforce Help Agent represents a significant step toward making AI-powered customer service easier to deploy, more action-oriented, and more closely connected to business outcomes.

By combining guided setup, built-in business knowledge, preconfigured actions, omnichannel deployment, and a pay-for-resolution pricing model, Salesforce is making it easier for businesses to adopt autonomous AI service agents without the complexity traditionally associated with implementation.

As AI continues to transform customer service, Salesforce’s outcome-based approach could help organizations focus less on how much AI they use and more on the real customer problems it successfully solves.