All Insights
A Guide to Rethinking Software Pricing for the Agentic AI Era
Pricing has never been easy to get right. More than a decade ago, a survey of 1,600 executives found that flawed pricing strategies were a leading reason that 72 percent of new products failed to meet their revenue targets.1 And subsequent studies have consistently reinforced those findings.
Agentic AI is making this more complicated than ever.
Agents change how buyers use technology and how they expect to pay for it. For enterprise software companies, the revenue opportunity is enormous — but we believe capturing it requires a fundamentally new approach to measuring and pricing value.
One of the first steps is evaluating whether your existing pricing metrics are built to last. Here are three questions every enterprise software company should be asking to get started.
Question One: Can Your Price Metric Keep Growing?
A good price metric should grow alongside the value a customer receives. For years, seat-based pricing did exactly that: software usage and value scaled with the number of users.
With Agentic AI, that formula can sometimes break down.
Agentic products can work continuously and autonomously, allowing output to scale without a corresponding increase in human users. A price metric tied to headcount, therefore, may put a ceiling on your ability to monetize your customers’ organic growth.
Consider a Vista portfolio company that prices its products by developer seats. The company observed that developer headcount was falling across its industry, particularly at junior levels. A survey of more than 200 of its own customers confirmed the concern: buyers expected their development teams to grow just 2-3 percent annually over the next three years.2
Stagnant headcount would limit the company’s ability to grow revenue, even as its products delivered more value. Its pricing metric had to change. With Vista’s support, the company is now transitioning to a model that captures the full economic value of its products.
This opportunity isn’t universal. If you sell software to an industry that is less technologically mature, your buyers may still expect healthy team growth in the years ahead. But for many software companies, seat-based pricing metrics may no longer provide the same runway for growth.
Question Two: Can Your Pricing Adapt to How Customers Use Your Product?
Traditional software usage was often relatively predictable: similar users performed similar activities and generated relatively similar levels of value. Agentic AI can create much wider variation in both product usage and customer value.
Pricing for that variability requires flexibility. Companies need time and evidence to understand which agentic capabilities customers adopt, how intensively they use them and which measures of activity best correspond to value before selecting a durable long-term price metric.
Consider a Vista portfolio company that is introducing agents across its platform. The company is working with customers to update its terms and conditions, and is introducing a flexible credit consumption model, so that its pricing model can evolve as usage patterns for its new products become clearer.
As usage data accumulates, the company can better understand how customers use its agentic products, where they derive value and which measures of consumption most closely track that value. It can then transition from credit-based pricing towards a model that more directly aligns with the value customers are receiving, for example qualified pipeline generated or meetings booked for its sales enablement agents, to directly align what customers pay with the value they receive.
Question Three: Could AI Reduce Consumption of Your Price Metric?
Seat-based pricing isn’t the only model ready for reinvention. Under certain usage-based models, agents are changing how customers consume – creating an opening to realign pricing with the greater value they deliver.
A leading provider of asset management software illustrates this dynamic. The company has historically priced based on tokens, which customers consume each time they open a module across its product suite. More activity meant more tokens consumed and more revenue generated.
Agentic AI is changing that. The same customers can now run sessions continuously, reducing the number of times they need to open individual modules. They get more from the product while consuming fewer tokens.
Every software company with a legacy monetization model — whether seat-based or usage-based — needs to ask whether its new agentic capabilities could reduce consumption of its pricing metrics. Those that realign their pricing can turn greater customer value into greater revenue.
Capturing the Pricing Opportunity
These three questions are primarily a diagnostic exercise. Across Vista’s portfolio, we find that most enterprise software companies identify at least one area where their existing pricing strategy needs to evolve.
From here, companies need to build a monetization strategy equipped to capture the value Agentic AI creates. That means determining what customers value, how to measure that value commercially and how to guide customers through the transition.
None of this is easy. Pricing agentic products may require companies to rethink commercial models that have worked for years. But the difficulty reflects the size of the opportunity. We believe Agentic AI can dramatically expand the value software can deliver and, with the right pricing model, the revenue it can generate.
Building a Pricing and Monetization Strategy for Agentic AI
