Salesforce’s AI Push Meets Reality Check

Analysts at KeyBanc Capital Markets are raising concerns about the adoption of Salesforce’s flagship AI platform, Agentforce, citing weak customer traction and product maturity issues. According to a recent research note, “Agentforce, as a product, just isn’t there” yet, based on customer checks and CIO surveys.

Adoption Challenges Mount

The KeyBanc report highlights several factors contributing to the slowdown:

  • Data readiness: Customers’ data often lacks the structure and quality needed for meaningful AI applications
  • Product maturity: The platform is still evolving, with three pricing model changes in roughly 18 months creating uncertainty for buyers
  • Outcome measurement: Enterprises struggle to connect AI interactions to tangible business results, making it difficult to justify investments

These concerns contrast with Salesforce’s sustained efforts to position Agentforce as its enterprise AI centerpiece, expanding it through new models, integrations, and deployment options.

Pricing Model Under Scrutiny

The move to consumption-based pricing has added another layer of complexity. Analysts note that:

  • It’s harder to budget for variable usage rather than fixed seat licenses
  • The lack of clear alignment between spend and outcomes creates uncertainty
  • Headless 360’s flexible architecture could make AI spending less predictable

“CIOs want a clearer line between spend and outcome, but that line isn’t clear enough yet,” said Bhupendra Chopra, chief revenue officer at Kanerika.

Data Foundation Remains Critical

Beyond pricing, the report underscores that Agentforce’s success hinges on enterprises modernizing their data infrastructure. As Gaurav Parab, principal research analyst at NelsonHall, explains:

“Most enterprises first want confidence that AI deployments will generate measurable business outcomes before committing to broader rollouts.”

This means many organizations need to invest in Data Cloud, integration, and governance initiatives before Agentforce can deliver reliable results at scale. As one example, a private equity fund administrator only saw consistent AI outputs after cleaning and structuring their data—demonstrating that data readiness is often the prerequisite for successful AI deployments.