Francesca Roche sits down with Kris Billmaier, EVP and General Manager, Agentforce Sales at Salesforce, to discuss what enterprises can learn from moving AI agents out of the demo environment and into real sales workflows.
As enterprises push AI agents from polished pilots into live sales environments, the real test is becoming clear: whether those agents can handle customer-facing work with enough context, accuracy, and trust to support real business outcomes.
Billmaier says Salesforce has been testing Agentforce Sales within its own sales organization, where the company found that over 60% of seller time was spent on non-selling activities.
That matters for customer experience because every hour lost to admin, prospecting, account planning, and manual follow-up is time sellers are not spending with customers.
The bigger lesson is that AI agents need to be treated like active members of the team, with training, supervision, and clear expectations from day one.
“Our biggest lesson early on was make sure your data is in good shape,” Billmaier says.
“[enterprises should] pick a use case, go deep, and treat the agent as if you would a junior employee that’s just getting started.”
One of the most interesting points is what happens when agents meet real customers, because live sales activity quickly exposes issues that never appear in a demo environment.
Billmaier explains that early email engagement from agents needed careful tuning before it reached the right quality, although Salesforce later saw agents “writing better emails, for example, than sellers were.”
That shift could change how sales teams spend their time, especially at the top of the funnel, where sellers often lose hours finding prospects, drafting outreach, and preparing early engagement.
For CX leaders, the data point is just as important as the productivity point, because AI agents can only make useful decisions when they understand the full customer story.
Billmaier puts it simply:
“Great AI comes from great data.”
He argues that CRM data alone is not enough, because “there are nuggets of gold in every sales conversation that happens,” whether those moments come from calls, emails, or other customer interactions.
Salesforce is using those signals to create summaries, insights, and coaching that feed back into the CRM, giving sellers more useful context before and after customer meetings.
Trust is another major theme, especially when AI agents begin influencing customer-facing sales activity.
“Trust is our number one value,” Billmaier says, explaining that Salesforce has built governance into the operating model through testing, quality assurance, live monitoring, permissions, and policy controls.
For enterprises still working out how to scale AI agents safely, the message is practical: start with the right data, choose one use case, define the goal, and prove the value before expanding.
