The software industry has entered one of its most consequential pricing and value inflection points since the shift to cloud computing. AI is not an incremental capability layered onto existing software; it is fundamentally redefining how software should create value and, in turn, how that value must be priced. The traditional model of seat-based subscription pricing was designed for systems that enabled human work, where scale was achieved by adding users. That model begins to break down as AI moves software from enabling work to executing it.
As AI assumes responsibility for tasks, decisions, and workflows, value is no longer
At the same time, organizations are reassessing a decade of software investments to determine whether they are positioned for an AI-driven and increasingly autonomous enterprise and operating model. This is not simply a technology upgrade; it is an architectural and operating model transformation driven by agentic AI, where systems can sense, decide, act, and learn across enterprise processes. This requires the agentic orchestration of AI agents that can interoperate across multiple software provider architectures and with other enterprises. Increasingly, this orchestration layer becomes the control point not only for execution, but for metering, governance and ultimately where economic value is captured.
However, many organizations are making premature assumptions about workforce reduction before establishing the foundational elements of AI readiness, including governance, orchestration and execution discipline. This creates a disconnect between ambition and capability, where cost actions outpace the ability to deliver sustainable value.
A central question emerging in this transition is what organizations should pay for AI-enabled software. For years, pricing has been anchored to human labor through seat-based models, creating a direct but now outdated linkage between users and value. AI breaks that relationship by introducing digital labor that operates independently of headcount. As a result, pricing models must evolve to reflect the value created by autonomous and semi-autonomous systems, as expressed through measurable units of work (e.g., transactions, decisions or automated process execution) rather than the number of licenses provisioned. At the same time, the scrutiny applied to professional services and labor costs is now being applied directly to enterprise software, forcing a broader reevaluation of what organizations are paying for and why.
Complicating this shift is the emergence of fragmented pricing approaches for AI-driven software. Many providers are combining platform-based pricing for AI agent operations with consumption-based models tied to tokens similar to those used in cloud computing, reflecting the economics of foundation models. While this hybrid approach is a natural starting point, token-based pricing anchors cost to technical consumption rather than business value, introducing uncertainty for enterprises that lack the ability to forecast usage in governed, production-scale environments. Tokens may measure activity, but they do not measure impact, introducing interest in pricing models that align more directly with outcomes.
This also introduces a new category of financial and operational risk that traditional IT cost management approaches were not designed to handle. Enterprises must now contend with:
In response, there is growing experimentation with value-based pricing models tied to specific business outcomes. While conceptually compelling, these models require a level of operational maturity that most organizations have not yet achieved. Defining, measuring and validating the value created by AI agents performing discrete units of work requires robust governance, clear attribution models and advanced analytics. Without these capabilities, outcome-based pricing risks becoming difficult to structure, negotiate and scale.
Importantly, outcome-based pricing is likely to remain selective rather than dominant. It is best suited to environments where work is highly measurable and attributable (e.g., transaction-intensive or process-driven domains) and has been historically tracked. For most enterprise software, the market will more likely converge on proxy value models, where pricing is aligned to units of work, transactions or decisions rather than fully outcome-based constructs. The shift to value-based models is therefore as much a measurement and governance challenge as it is a pricing innovation.
There is also a broader question of accountability. While the industry increasingly calls for software providers to align pricing with value, it is unrealistic to assign full responsibility for business outcomes to vendors alone. AI-infused software operates within enterprise-controlled environments shaped by automation and orchestration across workflows and processes tied to execution discipline. Value realization is inherently shared, and pricing models must reflect this reality rather than oversimplify accountability.
As these new variables are introduced, pricing is becoming more hybrid and complex, layering new models on top of legacy constructs. This complexity requires closer alignment between IT, procurement, finance and AI leadership to evaluate both the financial and operational implications of software investments. This is not simply alignment, but a shift in operating model:
To navigate this complexity, enterprises should adopt more structured pricing evaluation frameworks like autonomous level pricing (ALP) that can assess value based on autonomy and action, and support more intelligent decision-making in an AI-driven environment, including:
Despite claims that AI is dismantling the software as a service industry and pricing models, the reality is that consumption and value-based approaches were already emerging prior to the rise of generative AI. What AI has done is accelerate the timeline and increase the stakes. Software providers are now under pressure to ensure their pricing reflects the value delivered as digital labor becomes embedded in their platforms. At the same time, SaaS providers, hyperscalers and IaaS providers and emerging platforms are competing to control the economic model through which AI value is delivered, making pricing a central competitive battleground.
For enterprises, this means evaluating software not just on functionality, but on its ability to deliver measurable outcomes relative to existing operations and costs. AI is Software as I have articulated is and should be managed and governed accordingly.
Ultimately, pricing is becoming a strategic lever in the transition to the autonomous enterprise. Organizations must determine what constitutes fair value for software that is no longer just supporting work but performing it. This requires the ability to define, measure and govern digital labor in tangible units of work. In this new paradigm, both enterprise buyers and software providers must align on how value is defined, measured and sustained, as many of the pricing models that defined the SaaS era will not survive in their current form.
Regards,
Mark Smith