More than 50,000 Hours Back: What a Year of Manager Agent Taught Us

A year ago, Salesforce made a bet: Managers are the biggest driver of our agentic transformation. But their hands were full - data scattered across a dozen tools, administrative work crowding out coaching time, and a growing expectation to lead differently in this new agentic age.

So we built Manager Agent, an AI-powered partner that consolidates manager tools into a single entry point accessible in Slack. It handles the busywork of people management, like routine tasks and surfacing team insights, so managers can focus on the part of the job they love most: coaching their people. Manager Agent helps with tasks like surfacing HR information, resolving employee issues, providing coaching support, analyzing survey results, and facilitating performance evaluations and promotions.

We're not just making managers more efficient, we're helping them redesign how work gets done.

One year in, Salesforce managers hit 100% adoption of Manager Agent and have collectively reclaimed 57,000 hours. Here's what the journey taught us about managing in the AI era.

"Manager Agent is raising the bar for what great people leadership looks like. With a trusted partner handling the routine, managers can focus on what only they can do - know their people, grow their talent, and build high-performing, AI-fluent teams. We're not just making managers more efficient, we're helping them redesign how work gets done." - Pallavi Sebastian, SVP, Agentic Talent Experience

1. The biggest barrier wasn't skepticism - it was habit. Managers weren't resistant to using AI in their day-to-day work. They were busy. The shift happened when we stopped asking managers to go somewhere new and brought the tools into Slack, right into the flow of work.
2. Time savings matter less than what managers do with the time. More than50,000 hours is the headline. But the more interesting story is what managers did with the time back. Managers told us they used the recovered time for coaching conversations, thinking more carefully about team structure, and acting on engagement data instead of just collecting it.
3. Managers are our best drivers of AI fluency. When managers started using AI tools like Manager Agent, their teams followed. Our research found that employee adoption of internal AI tools climbed 32% when managers modeled the behavior. The best part? It happened organically - because employees experienced the benefits firsthand and felt safe to experiment.
4. We started with answers. Now we're giving insights. At launch, Manager Agent could answer simple HR questions, but the responses were generic. The same answer for every manager. Today, it knows the manager's team. It can surface information about their direct reports, feedback from stakeholders, and the places where their team may need support. That shift - from lookup tool to personalized coach - is where AI gets real.
5. The manager of the AI era looks less like an administrator and more like a coach. As admin work shifted to the agent, what surfaced was a clearer picture of what great management actually is: presence, judgment, relationships. And the managers who leaned into it thrived.

One year in, 57,000 hours back, and we're still learning. Time savings matter, but what matters more is what comes next: helping managers move beyond gut instinct to lead with clarity. The manager's role is evolving, but one thing is clear: Managers have never mattered more. So we're building with that in mind, putting people at the center of every decision, every feature, every step forward.