As AI agents begin to handle tasks once managed by employees, enterprise software companies face pressure to rethink seat-based pricing and adapt to a future where machines-not people-are the primary users of business platforms
For decades, enterprise software companies have relied on a simple business model: charge organizations for each employee who uses their platform. This seat-based pricing has been the foundation for giants like Salesforce and ServiceNow, whose revenues have grown alongside their customers' headcounts. But as AI agents become more capable of performing tasks that once required human intervention, this model is facing a fundamental challenge.
AI agents are now able to retrieve information, make decisions, and execute workflows across business systems. Instead of employees logging in to update records or move between applications, companies are increasingly deploying software agents to handle these repetitive or rules-based tasks. The shift raises a critical question for the industry: what happens to software revenue when the primary user is no longer a person, but an automated agent?
Platform Value Shifts Beneath the Surface
Some of the largest enterprise software providers are already adapting. Salesforce, for example, recently introduced Headless 360, an architecture that exposes its platform's capabilities through APIs and command-line tools. This allows AI agents to access data, workflows, and business logic directly, bypassing the traditional user interface. The move signals a broader industry trend: as screens become less important, the underlying systems-where business rules, permissions, and records reside-become more valuable.
According to reporting by TheStreet, platforms that have spent years building complex object models, permission structures, and audit trails are now positioned as essential infrastructure for AI-driven automation. These systems of record are difficult to replicate and provide the context AI agents need to operate safely and effectively. Yet, the transition is not without risk. AI agents may struggle to interpret legacy data structures or distinguish between similar fields, leading to errors that a human administrator would likely avoid.
Pricing Models Under Pressure
The most immediate impact for public software companies is on revenue. Seat-based pricing assumes that more users mean more value, but if AI agents can replace multiple employees, the link between usage and license count weakens. ServiceNow has already begun shifting toward consumption and outcome-based pricing, with more than half of its new business now tied to these models. Yet, defining what constitutes an "outcome" remains a challenge. Many vendors currently bill customers based on activity-such as the number of actions performed or AI credits consumed-rather than the actual business value delivered.
This approach can penalize customers for inefficiencies, such as poorly designed agents that repeat tasks unnecessarily. The companies most likely to succeed in this transition will be those that align pricing with tangible business results, like tickets resolved or cases closed, rather than raw activity metrics.
According to Salesforce's most recent quarterly report, subscription and support revenues reached $8.6 billion for the quarter ended April 30, 2024, up 11% year-over-year. However, analysts have begun to question how sustainable this growth will be as automation reduces the number of human users per customer account.
New Layers and Market Opportunities
The rise of AI agents is also creating opportunities for new companies that help bridge the gap between legacy systems and automated workflows. Most large organizations are unlikely to replace their core platforms simply because new AI capabilities exist-the cost and complexity of switching are too high. Instead, a new market is emerging for infrastructure that maps data, permissions, and workflows across existing environments, enabling AI agents to operate within established governance frameworks.
Few business processes are confined to a single application. Sales transactions, for example, often span customer relationship management (CRM) and enterprise resource planning (ERP) systems. AI agents can read data across these platforms, but acting on it requires careful coordination of permissions and compliance rules, which rarely align perfectly between systems.
Adoption Pace and Investor Risks
For investors, the pace of this transition is a key concern. While the conversation around AI agents suggests rapid disruption, the reality is likely to be slower, especially among large enterprises with deeply customized systems. Trust and reliability are paramount in this market, and organizations are cautious about overhauling mission-critical infrastructure based solely on new technology capabilities.
Smaller businesses, with less customization and lower compliance burdens, are often the first to adopt agent-driven models. Some are already running operations on lightweight platforms with minimal legacy infrastructure. Broader enterprise adoption may take five to ten years, as companies gradually renegotiate contracts and adapt to new pricing structures. In the near term, software vendors face pricing pressure as customers push back on seat-based models that no longer reflect actual usage or value.
As AI agents become more central to business operations, the competitive advantage of enterprise software may shift from user-friendly interfaces to robust, well-governed back-end systems. Companies that built their moats around powerful UIs will need to prove their value in a world where the interface is no longer the main point of contact.
Enterprise software pricing is at a crossroads as automation accelerates. The next few years will test whether established vendors can adapt their business models to a future where machines-not people-are the primary users of their platforms.
Enterprise software platforms are often built around complex permission models, audit trails, and business logic that have evolved over years of real-world use. These features are critical for compliance, security, and operational reliability, especially in regulated industries. As AI agents take on more tasks, companies must ensure that automation respects these controls and does not introduce new risks. For organizations considering a shift to agent-driven workflows, it is essential to evaluate not just the potential efficiency gains, but also the readiness of their underlying systems to support safe and effective automation.
