Agent AI - an emerging packaging pattern
Steven Forth is CEO of Ibbaka. Connect on LinkedIn
Ibbaka recently released its annual survey on AI monetization. As part of the survey design process we had to update some of the core frameworks that Ibbaka uses to understand value, pricing and monetization. The three key areas that we updated are
Packaging patterns (this post)
Growth motions (in preparation)
Pricing metrics (in preparation)
Please take the AI Monetization in 2025 Survey
How are packaging patterns changing?
In our research into Net Revenue Retention (just published, you can get the report here) we found new patterns emerging in B2B SaaS driven by the rapid emergence of second generation generative AI applications. This transformation is still early days, but it is already making certain packaging patterns more prominent, especially those based on AI agents.
Ibbaka has added two patterns to its common SaaS packaging patterns: Agents and Family of Agents (on the right above).
There were some write-in votes for these patterns in the 2024 NRR survey referenced above. More important ares Salesforce’s ambitions around agents (to see one billion deployed by the end of 2025) and the support it has been rolling out through Agentforce. Salesforce played a major role in creating the SaaS or subscription economy. Will it play a similar role in the monetization of B2B generative AI? It does ont have to do this directly. Its role a an example can be transformative. If Salesforce decides to package AI in agents many will follow.
The AI Agent pattern
Agents have been part of the B2B software landscape for many years, sometimes going under different names. Many Robotic Process Automation (RPA) applications are examples of simple agents. Grammarly began as an agent to improve writing and is now a family of agents to generate content and manage conversations in many different contexts (though it is still priced as a single agent). The various generative AI powered co-pilots and assistants that are cropping up everywhere could be thought of as general purpose agents.
In B2B software the most effective agents, the ones that are getting the most commercial acceptance and that are leading pricing innovation are agents that have a specific, and generally limited, set of functionality that can be clearly mapped to a task done by a specific business function. Two of the most common examples are in sales and in customer support.
Sales agents include AI powered Sales Development Representatives that are rapidly replacing cold emailing and cold calling and that are invading our social networks. These agents have a very specific goal. Generate a qualified lead, gather information about that lead, and set up an initial conversation. These agents are outperforming humans and in most B2B software the SDR role will have evaporated by the end of 2025.
Customer success agents are also very popular. Intercom, Zendesk, a lot of what Salesforce is proposing, are customer success agents. Here the current pricing metric is ‘customer tickets successfully resolved.’ The tricky part here is what ‘successfully resolved’ really means. Right now it tends to mean ‘the support ticket was closed’ (the same is true when a human is providing the support). Down the road, it will come to mean ‘the customer renewed’ or ‘the customer made an additional purchase.’
Many other areas are open to AI agents. Anytime there are clearly defined tasks that a business function needs to complete, humans will be replaced by agents.
The dominance of agents will finally put an end to user based pricing. With agents the number of people employed for a task will collapse if not disappear altogether. User based pricing of an AI Agent for Customer Success make no sense. AI Agents will have a forcing effect in getting B2B SaaS companies to move off per user pricing models.
Agent pricing will be based on a combination of
The number of agents
The complexity of the task the agent performs
The value of the task
The performance of the agent
The outcome of completing the task
The Family of Agents pattern
The power of B2B AI agents is that they are designed to perform a specific task inside a business function or process. This focus makes it easier to design and train agents and to measure their performance. But most significant business processes require the completion of multiple tasks. Agents can be extended to do more tasks, but that will often lead to higher development costs and lower performance. The solution will be families of agents that can be connected to gether to cover off omre of the business process.
How will these families of agents be priced? Will there be a discount for buying more members of the family or will some companies try to charge a premium as the family delivers more value than any combination of agents working separately? This is sometimes know as the ‘team premium.’
To make matters more complicated, down the road one will likely be able to buy agents from different companies and orchestrate their behavior yourself or with a third-party agent orchestration platform. In fact, Salesforce is positioning Agentforce to be that orchestration platform, though it is likelty to have a lot of competition in a rapidly evolving market.
Will AI Agents become a dominant pattern?
We reached out to our community with a simple poll to ask people what they thought about this question.
“What role will AI agents play in packaging and pricing B2B functionality?”
Become the dominant paradigm
Be one of several patterns
Be an edge case
Be of marginal use
The poll was shared across our personal feeds, the Design Thinking Group, the Professional Pricing Society, and the Software as a Service - SaaS Group. It is of course a small and biased sample, but suggestive.
From an Ibbaka LinkedIn poll open the first two weeks of November 2024. N = 64
Most people AI Agents to be one of several common patterns. That is the safe bet. A smaller number of people thought it could become the dominant paradigm, which would indeed be interesting. Having a dominant paradigm can greatly accelerate adoption and remove uncertainty, which is likely what Salesforce is aiming for. And of course a significant number of people, more than 23%, think this is overblown. We shall see.
Will Salesforce be the dominant player? It is going to have competition from Google and the OpenAI Microsoft team. Expect a major commercial announcement from Microsoft in the next few months. Agents are a threat to its productivity suite franchise. There are also upstarts entering the market: Aisera (the agentic AI copilot), Agent.ai (the professional network for AI agents) and Novus (AI orchestration for Enterprises) that could disrupt established players.
Ibbaka expects the AI Agent pattern to be the dominant, but not the only pattern for monetizing B2B AI.
More on Generative Pricing
Contact Ibbaka today and discover how our unique B2B SaaS solution can increase sales, improve retention, and drive growth.