Salesforce is fully committed to agentic AI as a new growth engine. The company is emphatically positioning its Agentforce platform as a reinforcement of the SaaS model, at a time when the market is discussing declining growth and falling valuations in the SaaS sector.
This became apparent during the presentation of fourth-quarter results, in which Salesforce CEO Marc Benioff stated that AI agents fundamentally improve SaaS services rather than undermine them. According to Benioff, agents-as-a-service create new value for customers and increase the appeal of existing SaaS platforms.
During the earnings call, Salesforce announced that it has attracted more than 180 customers for Agentforce IT Service since October, according to The Register. This offering focuses on IT Service Management and is seen by analysts as a direct competitor to ServiceNow.
Salesforce indicated that several organizations have switched from ServiceNow to Salesforce for IT services. This move puts the company firmly in an enterprise market where ServiceNow traditionally has a strong position.
Mixed reception for financial results
For the full fiscal year, Salesforce achieved revenue of $41.5 billion, exceeding its previously issued forecast of up to $40.9 billion. A significant portion of that growth came from the acquisition of Informatica, which contributed $399 million to revenue. As a result, organic growth remained more limited.
Revenue in the fourth quarter came in at $11.2 billion, an increase of 12 percent year-on-year. At the same time, remaining performance obligations were lower than analysts had expected. Despite a dividend increase and an announced $50 billion share buyback program, Salesforce’s share price fell 5.6 percent after trading hours.
New measurement method for AI agents
Salesforce also introduced a new metric to provide insight into AI agent usage: Agent Work Units. This metric measures how many concrete tasks AI agents perform within production environments such as Agentforce and Slack.
According to Salesforce, Agent Work Units should help organizations better quantify the business value of generative AI by measuring not only token usage, but also actual output, such as resolving customer queries or automating workflows.