Humans are often hired with domain expertise and can infer business logic. AI agents, however, are like "newborn children"; they only know what they are explicitly taught through data. To make an agent understand a simple metric like "pipeline," you must provide extensive metadata and context that a human would already know.
Moving data between warehouses is costly and often unnecessary. The industry is shifting to a "zero-copy" model where data is accessed and activated where it lives. This allows companies to tap into existing data investments without redoing integrations, focusing budgets on value creation instead of data movement.
The value of high-quality data has always been abstract. Now, with AI agents, it's quantifiable. You can directly measure an agent's performance—its accuracy, efficiency, and cost—when fed clean, well-structured data versus poor data. This provides a clear business case for data governance and hygiene investments.
The role of a Chief Data Officer is shifting. Beyond serving executives and departments with metrics, data teams must now prepare and structure data specifically for AI agents, which are becoming key "workers" and data consumers within the enterprise, effectively becoming a new customer base.
Contrary to fears of job displacement, AI agents are voracious data consumers needing far more context than humans. Salesforce's CDO finds this dramatically increases the workload and hiring needs for data teams, as they must produce a much higher volume of trusted, agent-ready data to fuel the new automated workforce.
Salesforce's internal data team operates within the product organization, acting as the first and most demanding user of its own software like Data360. This "Customer Zero" approach ensures products are battle-tested at massive scale, turning internal operational challenges into R&D opportunities for their customers.