Salesforce has introduced a new way of pricing Agentforce: pay-per-resolution. The company yesterday revealed a new Agentforce Help Agent – an autonomous service agent designed to help enterprises resolve customer issues “from start to finish, on day one”.
In what will likely come as a welcome change for Salesforce customers, Salesforce explained its new pay-per-resolution pricing model, with businesses only charged when Help Agent resolves an issue autonomously. “If the customer asks for a human or walks away unhappy, there is no charge,” Salesforce says. “Your cost is tied to outcomes, not activity.”
This is a huge shift in the way Salesforce charges its customers – and a far cry from its traditional seat-based pricing model. Let’s take a look at what this means.
Agentforce Help Agent’s Pay-Per-Resolution Explained
Salesforce says that organizations using the new Agentforce Help Agent only pay when it autonomously resolves an issue from start to finish.
If a customer provides negative feedback or asks for human escalation, there is no charge, and the agent passes the customer context to the service team.
Data 360 and Agentforce are both unmetered during the agent interaction.
“This means companies don’t have to worry about forecasting consumption or overages, and their investment is tied directly to successful customer resolutions,” Salesforce said in the announcement.
Pay-per-resolution pricing will be available July ’26, Salesforce says.
A Massive Shift in Pricing Model
Agentforce pricing has been a topic of discussion in the community for some time, and Salesforce has been taking notice.
Agentforce was first announced at Dreamforce ‘24, and the consistent messaging from Salesforce about how much it costs in those early days was “$2 per conversation”.
With an early model of usage-based pricing, Timo Kovala, Marketing Architect at Capgemini, wrote in his SF Ben article about common Agentforce pain points: “One of the main gripes with Agentforce has been its “$2 per conversation” pricing model. With this model, Salesforce sought simplicity and ease of communication, but ended up limiting flexibility and obscuring costs.”
There had been some confusion about how exactly certain scenarios would play out, and whether businesses would have to pay for Agentforce interactions that were not creating value in any real way. Even the concept of consumption-based pricing was fairly new to Salesforce at the time, though not unprecedented – Data 360 (formerly Data Cloud) uses this.
In any case, the move away from traditional seat-based pricing by the SaaS company was an adjustment.
We wrote about why Agentforce adoption was slower than expected in April 2025, and Ian Gotts, CEO and Founder of Elements.cloud, told us at the time: “They don’t know how to price them yet, and right now, nobody wants to hand out a blank cheque – they want caps and predictable ROI.”
But in May that year, Salesforce announced that it was introducing “pricing innovations” including Flex Credits – a consumption-based model meant to align cost with outcomes – along with a Flex Agreement and new Agentforce user licenses. It seemed apparent at this time that the company was addressing hesitancy from the community around adopting a consumption-based pricing model for something where outcomes were not guaranteed.
A week after the pricing update, we covered how the ecosystem reacted to the new pricing model, and we found it to be largely positive.
Though Chief Executive Officer at Attentis Consulting, Steve Snell, had rather humorously compared the pricing schema to his “1991 cell phone data plan where every month’s bill was an adventure into the great unknown”.
This move from seat-based pricing to consumption-based pricing and now to outcome-based pricing is an interesting shift from Salesforce. This new shift seems to imply that the SaaS company is not only listening to customers’ pain points and barriers to adoption for Agentforce, but is also confident enough that Agentforce will be able to successfully resolve customer problems, to the point of this being profitable for them.
Agentforce Help Agent
Salesforce says that its new Help Agent comes with a library of available actions that can run on real workflows to manage cases, schedule appointments, and update orders.
The company said it works “everywhere your customers are”, across voice, web, portal, and messaging.
“We know it works because we have run it ourselves: on help.salesforce.com, Agentforce has handled 4.3 million inquiries and resolved 70 percent of them, and everything we learned doing that is built in,” the company said in its announcement.
If this new Help Agent outcome-based pricing creates traction for Salesforce, it would be a fine demonstration of the value of “dogfooding” in the agentic age. It would also be a huge return on investment for the cost that Salesforce has paid in the initial rocky rollout of their own help agent in 2025.
Agentforce Help Agent and Agentforce Customer Service Portal are generally available July ’26.
Final Thoughts
There are challenges to implementing outcome-based pricing for customer satisfaction because Salesforce incurs costs for every token generated by their models, or at a minimum, the cost for the associated compute power.
As SF Ben Editor-in-Chief Peter Chittum explains: “For this pricing structure to be viable, token and compute costs must be sufficiently low to protect Salesforce’s financial health, and the AI model must maintain a high success rate to prevent users from consuming excessive tokens without achieving a resolution.”
But there is also a vulnerability of pay-per-resolution models to bad actors who could potentially abuse the system. A user could repeatedly interact with a help agent and claim dissatisfaction, intentionally driving up token costs for Salesforce without achieving a successful outcome. Salesforce will be aware of this, so adopting this pricing model implies that the company is either assuming significant financial risk – or has a great deal of confidence in the model’s effectiveness, and plans to mitigate any such risks.
