What's Happening?
Salesforce is developing new pricing structures for its AI products, moving away from traditional per-user licensing. This shift is driven by the increasing use of AI agents and APIs, which complicate the existing model designed for human users. Bill
Patterson, executive vice president and general manager of CRM applications, indicated that Salesforce is devising a pricing structure that aligns with the benefits customers realize from AI technology. One proposed model involves charging for AI customer service agents based on the number of cases they resolve. However, outcome-based pricing presents challenges for agents operating across multiple disciplines, making it difficult to identify a single measurable result. Consequently, Salesforce is also exploring bundles and Flex Credits, which charge customers based on consumption. Mike Spencer, Salesforce's deputy CFO and head of finance, acknowledged that license revenue is currently hindering growth, and the company is experimenting with various pricing structures and contract frameworks, including conventional seat licenses, consumption-based Flex Credits, and outcome-based fees. The company anticipates that consumption revenue will become a significant portion of its earnings within three to five years as more customers implement AI systems.
Why It's Important?
This strategic pivot in Salesforce's pricing model reflects a broader industry trend where the value proposition of software is shifting from user access to tangible outcomes and consumption. For U.S. businesses, particularly Fortune 500 companies heavily reliant on Salesforce, this change could significantly alter their operational budgets and how they evaluate technology investments. The move towards outcome-based pricing could incentivize businesses to focus more on measurable results from their AI deployments, potentially leading to more efficient and impactful use of AI. However, it also introduces complexities in cost prediction and contract negotiation, as highlighted by Gartner's warning to Salesforce users about the potential disappearance of all-you-can-eat agreements. The transition could benefit companies that can clearly define and measure the outcomes of AI integration, while those with less defined metrics might face challenges in optimizing their spending. This evolution in pricing models could set a precedent for other enterprise software providers, influencing how AI services are valued and adopted across various U.S. industries.
What's Next?
Salesforce will continue to refine and implement its new pricing models, with consumption revenue expected to become a material share of its earnings within the next three to five years. The company is likely to engage in ongoing discussions with customers to determine objective measures for outcome-based pricing, which Spencer noted is a key challenge. Customers will need to adapt to these evolving contract structures, potentially moving from traditional annual commitments to models like Agentic Enterprise License Agreements (AELAs) or Salesforce Commit, which resemble hyperscaler models with upfront spending commitments. Preet Takkar, PwC's global and US Salesforce leader, anticipates that outcome-based pricing will become far more prevalent by 2030, suggesting a gradual shift away from capped agreements towards spending commitments and eventually charges tied to results. Businesses should proactively engage with Salesforce to understand the economic implications of these changes and negotiate terms that align with their specific AI adoption strategies and desired business outcomes.
Beyond the Headlines
The shift in Salesforce's pricing strategy underscores a fundamental re-evaluation of value in the age of artificial intelligence. Beyond mere cost, this move highlights the increasing importance of demonstrating concrete business outcomes from technology investments. It challenges the traditional software licensing paradigm, where value was often tied to the number of users, and instead emphasizes the impact of AI on productivity, efficiency, and problem-solving. This could lead to a more performance-driven relationship between software vendors and their clients, fostering greater accountability for the return on investment in AI. Furthermore, the complexity of defining and measuring outcomes for AI agents operating across diverse functions raises ethical and methodological questions about how to fairly attribute value and cost. This evolution could also accelerate the development of more sophisticated analytics and reporting tools to track AI performance, ultimately shaping how businesses perceive and integrate AI into their core operations.
