The software industry has long relied on the seat based pricing model. For years, this approach provided a straightforward path to revenue. It did not always require the software to demonstrate its full impact on an organisation's overall performance. If the product was useful for daily tasks, companies purchased it. This model allowed vendors to focus on user counts rather than building complex systems to quantify the holistic value they delivered. Attribution of value was often a secondary concern, as the incentive to track specific outcomes was less immediate.
That era is now evolving.
The shift toward usage based pricing was a significant step, but it introduced its own complexities. With variable costs, customers began paying closer attention to their spending. They are now more actively evaluating whether the usage justifies the expense. Return on investment (ROI) becomes a central question when budgets are not fixed. Finance teams, who typically prefer predictability, can find variable spend challenging, leading them to ask more pointed questions about the value they receive in return.
The rise of AI agents is accelerating this shift. This technology changes the traditional math. If an AI agent can automate tasks that previously required multiple human employees, the number of human operated seats will naturally decrease. This presents a challenge for software companies facing potential pressure on top line revenue from traditional seat contracts and margins due to rising compute costs. The underlying issue with both seat based and usage based models is that they can avoid the direct conversation about the value delivered. A disconnect can arise when the price does not clearly align with the tangible results a customer achieves.
This has led to growing interest in outcome based pricing.
Consider the advertising industry. Performance marketing has established clear metrics for what an action, like a click, is worth. This creates a direct link between spend and result. The software industry is beginning to explore similar territory. For example, Intercom's pricing for Fin.ai, based on per resolution, is an attempt to align vendor incentives with customer success. In this model, the customer pays only when a defined outcome, a ticket resolved, is achieved.
This model is not a fit for every product. For a tool like Slack, which enhances productivity in diffuse ways, quantifying a single, tangible outcome is complex. In such cases, the focus may shift to identifying proxy metrics that correlate with improvement, a "north star" indicator of value.
As software pricing moves closer to measurable outcomes, the next challenge is building the infrastructure required to track, attribute, and monetise that value at scale.
Building the infrastructure
Recognizing value is only the first step. Capturing it requires a fundamentally different kind of infrastructure.
The discipline of sophisticated monetisation has, for some time, been more advanced in other sectors. Internet companies and marketplaces have spent years perfecting it. Meta and Google developed complex systems for ad monetisation. Uber and Amazon refined their marketplace models. The software industry, by contrast, often relied on the simplicity of the seat model. As that model faces new pressures, software companies are now looking to build the kinds of attribution systems that ad tech companies have long utilised.
This is a complex but necessary undertaking.
It involves creating internal systems that can track and connect product usage to the delivery of value. Many software organisations are not yet equipped for this. They may lack the infrastructure to know which specific features or usage patterns correlate with customer renewals or success. Ad companies, with their need to prove the value of an impression, have had to build robust attribution capabilities. The software industry can learn from their approach by investing in data infrastructure that links product actions to financial outcomes.
This shift is already creating new demand for specific talent.
The most valuable skill sets are emerging at the intersection of data engineering, product analytics, and monetisation strategy. We can expect to see a growing need for data professionals who understand business metrics. The role is not just about moving data but about interpreting it in the context of pricing models and building automation that supports product experimentation. For many companies, this will become a budget priority.
There is a broader strategic consideration here beyond just hitting revenue targets.
The software market has historically been fragmented, with room for many players. As pricing models, a key differentiator, come under pressure, the market could see consolidation. AI agents are highly scalable, and the companies that successfully integrate them will have an advantage, particularly if they also crack the code on monetisation. Building a deep understanding of attribution and value based pricing can create a significant competitive edge. Those who succeed in both areas could capture a larger share of the market for a given utility.
These efforts are interdependent. It is difficult to build compelling AI features in isolation without a parallel plan for how to capture the value they create. The most successful companies will likely be those that treat pricing as an integral product feature, developed in tandem with the technology itself. The tools and technologies are available. The challenge for leadership is choosing to evolve established systems proactively rather than waiting for external pressures to force the change.
Author's Note: I wrote the initial draft of this article based on my perspectives, and used AI to structure and proofread the final version.