Look what the Easter Bunny brought Hubspot customers - a shift to outcome-based pricing for Customer and Prospect Agents

HubSpot has announced that it is switching two of its Breeze AI agents: Customer Agent and Prospect Agent to an outcome-based pricing model. Why these AI agents and why the change?

With outcome-based pricing, the customer only pays when the agent achieves a defined outcome, such as resolving a support issue, qualifying a lead or booking a meeting, or completing a defined job start-to-finish.

HubSpot’s change in agent pricing

Today, HubSpot announces that its Breeze Customer Agent and Breeze Prospect Agent will now be charged using an outcome-based pricing model starting April 14.

According to HubSpot, companies are facing increasing pressure to justify agentic AI spend and prove ROI, and one of the best ways to help them is to charge based on outcomes. The company believes that AI should be measured in outcomes not output, but most agents aren’t priced this way because they can’t promise consistent results.

In HubSpot’s case:

• Customer Agent: When the AI agent resolves a conversation without the need to assign to a human.
• Prospect Agent: When a qualified lead is passed on to sales.

One of the primary reasons HubSpot is able to shift these two agents to a pricing model based on task completion is because the agents are built into the HubSpot platform where they have access to the customer and business context needed to deliver reliable performance. For example, these agents can access contact data, relationship history, and other business information.

The Customer Agent delivers a 65% conversion rate and 39% faster resolution times across 8,000 plus HubSpot customers. This pricing change means customers shift from paying $1 per conversation to $0.50 per resolved conversation.

With the Breeze Prospecting Agent, HubSpot says that customer activations are up 57% quarter over quarter. It moves from a monthly recurring charge for every contact enrolled to $1 per lead recommended for outreach (so you only pay when a prospect is qualified and handed over to the sales team).

HubSpot says this pricing change allows customers to move faster, experiment more, and trust that spend is tied to real results.

The different approaches to agent pricing

As AI agents continue to grow in use among organizations of all sizes, pricing has become an interesting conversation because no one has figured out how to do it consistently. Research from paid.ai found that 75% of vendors adding AI agents have no systematic approach to pricing them.

A few approaches include:

• Usage-based (45%)
• Per agent (20%) - this is basically a full-time-employee (FTE) replacement
• Workflow-based (15%) - Salesforce Agentforce is one example.
• Outcome-based (20%) - Intercom’s Fin is an example

The research notes that outcome-based pricing yields the highest margins with low churn, while usage-based pricing is the most vulnerable to commoditization.

One of the first, but not the only one

Although HubSpot is one of the first software vendors to shift at least one of its agents to an outcome-based pricing model, it isn’t the only one.

Intercom’s Fin AI agent uses outcome-based pricing, charging $0.99 per outcome. It defines two types:

1. Resolutions (similar to HubSpot) where the customer confirms their issue was resolved or they don’t request more help after Fin responds.
2. Procedure handoff where Fin successfully executes a procedure that is configured to end in a handoff to a human agent (e.g. subscriptions or refunds).

Zendesk is another company that moved to outcome-based pricing to align its costs with value for the customer.

According to the paid.ai research, most customer support agents are moving to outcome-based, while only 20% of RevTech (AI SDRs and AI AEs and Marketing Agents) are following this pricing model. The research notes that one AI SDR company charges $200 per attended meeting and sees 94% gross margins.

Then there are vendors who charge using a hybrid model. For example, Intercom includes Fin AI Agent with its helpdesk, which it charges per seat. So, you have that combination of outcome-based when Fin is engaging with customers, and the per seat cost for help desk agents.

Of course, HubSpot is similar in that you still pay for the HubSpot platform, then pay for the use of the AI agents within the platform. Other agents in the HubSpot ecosystem still work on a credit model where you buy credits to consume agents and other AI features.

Defining success and monitoring results are key

Outcome-based pricing for AI Agents is an emerging model that makes sense in some situations. If the vendor and customer can work together and clearly define what “success” looks like, then outcomes can work. It’s also essential that the vendor providing the agent can directly influence the results from the agent. If they have no control or influence, then the vendor can’t be sure if the agent is providing the results needed.

Businesses that use AI agents with outcome-based pricing need to ensure they know how success is measured and that the agent is auditable to prove the outcomes were truly achieved.

And while outcome-based pricing makes sense for agents like the Customer Agent and Prospect Agent, someone on the customer side needs to be responsible for monitoring the results to ensure the billing is accurate, which can possibly lead to more effort than when working with a different pricing model.

My take

HubSpot is definitely taking the right approach to pricing these two agents going forward. It will be much easier for customers to prove the ROI of using these agents. Proving the ROI of other agents is trickier because there is no way to define clear outcomes (the outcomes vary based on what the agent does).

However, you can see how it may become challenging for both HubSpot and customers to track each agent they use, how they are paying for it, and how they define and track outcomes.