For most of software’s history, pricing reflected how it was built and used. It was predictable, and the software was purchased as a tool. Perpetual licenses and, later, “seat-based” subscriptions were logical models for years.
With AI-enabled software continuing to accelerate its growth and usage, that way of thinking doesn’t work anymore.
The move to cloud computing introduced consumption-based pricing aligned to usage. As software becomes more adaptive, autonomous and capable of driving outcomes, pricing models tied to access or activity are starting to feel outdated. Software has changed, and pricing should, too.
AI-enabled software is fundamentally different from traditional enterprise software. It can reason, take action and adapt in real time while consuming compute. That’s a shift from traditional software, which delivers value through dashboards and predefined workflows.
Trying to price that new type of software using old-school static thresholds or fixed constructs is a mismatch from the start.
Moving beyond comforts of subscription pricing
Subscription pricing has stuck around because it feels safe. Organizations can budget for it, and vendors benefit from steady recurring revenue. Usage-based pricing aligns cost with consumption, especially in infrastructure and developer-focused platforms.
Metrics such as tokens, credits or compute units don’t measure outcomes. They’re essentially proxies for value; not value itself.
Two organizations can consume roughly the same amount of AI resources and see dramatically different business outcomes. Treating those scenarios as equivalent doesn’t make sense.
In fairness, AI introduces uncertainty on both sides of the table. Vendors face variable infrastructure costs driven by inference and compute demand. Buyers struggle to forecast spending when usage fluctuates, and value shows up unevenly across teams and use cases.
Hybrid models that blend subscriptions with usage commitments or AI credits can help manage complexity and serve as an interim approach.
Hybrid models in action
Salesforce has taken a hybrid approach with Agentforce, introducing a bundled model that prices AI based on the actions it performs, like executing workflow updates or modifying records. The result combines seat-based access with consumption signals, moving away from seat counts as the only value driver.
Adobe also shows how pricing can evolve with value. While its Creative Cloud product still prices access per user, newer AI features use usage-based credits, with customers paying more as they generate more output. It’s a practical hybrid model that preserves subscription stability while moving beyond seats alone.
Acknowledging software’s more active role
Software is no longer just something you buy and deploy. As software becomes an active participant in day-to-day operations, you’re essentially hiring digital teammates.
While work like resolving customer inquiries and optimizing workflows remain human-led, AI-based software is taking on real responsibility for outcomes. Pricing based purely on access starts to feel disconnected from the greater role software now plays.
Performance already drives compensation in other areas of the organization: Sales teams are paid based on results, and service providers are paid for outcomes delivered. AI makes it possible to extend that same logic to software.
Value-based pricing aligns incentives more cleanly. Vendors are rewarded for delivering measurable business impact, not for encouraging more usage. Customers pay for results that matter instead of abstract activity measures.
The operational roadblocks for value-based pricing
If value-based pricing makes so much sense, why hasn’t it been stronger out of the gate? The hesitation is less philosophical than it is operational.
Defining meaningful outcomes requires alignment across the business, IT and procurement. Measuring those outcomes demands the right data, analytics and agreement on how value should be attributed. Translating impact into commercial terms pushes sales, finance and legal teams into unfamiliar territory.
In fast-moving environments, speed and simplicity often win. Usage-based pricing is familiar, relatively easy to implement and quick to bring to market. In AI-driven spaces, where innovation cycles are short and expectations are high, that familiarity is appealing. The tradeoff is that it delays the inevitable.
Three steps to prepare for value-based pricing
Organizations shouldn’t wait for perfect outcome-based models before preparing for them. There are practical steps that can be taken today:
-
Start measuring outcomes. Even if contracts are still usage- or credit-based, teams can track the metrics AI solutions are meant to influence. Productivity, revenue impact, risk reduction and customer experience all provide helpful insight into how value is being created.
-
Experiment with hybrid structures. Introducing outcome-linked elements into traditional agreements lets vendors and customers learn without taking on excessive risk. Over time, these models will build trust and transparency.
-
Expand AI literacy beyond IT. Procurement, finance and business leaders need a shared understanding of how AI creates value, in probabilistic (not deterministic) ways. That fluency makes outcome-oriented pricing much easier to govern.
Embrace the inevitability
The software and platforms industry will continue to test and refine pricing approaches. Some will scale, and others won’t. Differences across industries and use cases will persist.
Value-based pricing is an inevitable reality as AI transforms software from a passive tool into an active contributor to business performance, leading to pricing that will increasingly reflect outcomes rather than inputs.
