Ep 323 – Brandon Roberts (SVP Talent Strategy & Workforce Transformation, ServiceNow)

1. Start with the work — not the AI tool

One of the first things ServiceNow did was map every role in the company and ask a simple question: What does this person actually spend their time doing?

From there, they assessed which tasks were most likely to be augmented by AI. The result was an “AI heat map” that showed where work was likely to change the most — and where HR should focus its energy first.

The important part: the heat map wasn’t treated as the answer. It was an input into workforce strategy.

“That heat map is not a solution, it’s an input to a strategy.”

For People teams, this is a much more useful starting point than asking, “Which AI tools should we buy?” Break roles into tasks, identify where the work is changing, and then use that signal to decide where you need new skills, redesigned roles, or an entirely different operating model.

2. Productivity gains should trigger role redesign — not just cost cutting

ServiceNow saw this firsthand inside HR shared services.

As AI got better at answering routine employee questions around benefits, policies, manager changes, expenses, and other requests, one shared services employee could support far more employees than before. Brandon said the ratio went from roughly 400 employees supported per person to around 1,000.

But they didn’t stop at the productivity number.

“We didn’t just squeeze productivity. We didn’t cut costs. What we decided was we were going to change their role.”

Some employees were reinvested into improving the AI itself — maintaining better data, understanding how the technology worked, and making sure employee questions were answered correctly. Others were upskilled into more advanced HR roles.

That’s the bigger opportunity for HR: when AI removes work, don’t just ask how many people you need. Ask what higher-value work those people could now be doing.

3. Build a system for spotting where work is changing

The AI heat map is only one signal ServiceNow uses.

They also look at actual technology adoption through an AI control tower — essentially a view into which AI tools and use cases are being used across the company, by whom, and how.

Then they combine those signals with business context.

“You use the technology as a signal, you take the work as a signal, and then we bring together HR business partners, someone with experience in role redesign and workforce planning, and then AI transformation experts.”

Those cross-functional groups can then go into a business and ask much better questions:

1. Do we need a new role?
2. Do we need a different org structure?
3. Does the operating model need to change?
4. Are the skills we hire for today still the skills we’ll need tomorrow?

The takeaway: build mechanisms that help HR identify where transformation needs to happen before the org chart becomes obviously outdated.​

4. Give HRBPs agents that let them answer business questions in the room

Brandon shared an example of an HRBP sitting in a leadership offsite where the business leader raised a concern: the organization felt too complex and too slow.

Historically, that HRBP might leave the room with an action item, spend days meeting with people analytics, employee listening, research, and other teams, and then come back with an answer.

ServiceNow is trying to compress that cycle dramatically.

They’ve built agents that can pull organizational data, analyze spans and layers, summarize employee survey feedback, and surface relevant insights while the conversation is still happening.

“Our view is that a lot of what AI agents are allowing now is that you can do this in the room and get this data in real time.”

That changes the HRBP role. The value is no longer in collecting the data or coordinating the handoffs. The value moves toward interpretation, judgment, recommendation, and influence.

5. AI makes judgment more important, not less

Giving HR faster access to insights does not automatically create better decisions.

Brandon repeatedly came back to judgment as one of the most important skills in an AI-enabled organization. An agent can surface the data. It can summarize what employees are saying. It can help generate analysis.

But someone still has to decide what the information means — and what the organization should do about it.

“You can get data and get the wrong insight.”

That means reskilling can’t stop at teaching employees how to prompt a model or build an agent.

For HRBPs specifically, the development opportunity is learning how to turn information into a meaningful strategy, challenge assumptions, understand the business context, and influence leaders toward the right action.

The more operational work AI absorbs, the more valuable those human skills become.​

6. The future HR org may have far fewer archetypes

Brandon sees the traditional HR operating model simplifying over time.

His current thinking centers around three broad profiles: people strategists, people technologists, and people scientists.

People strategists are the evolution of the HRBP — deeply connected to the business and helping leaders transform the organization. People technologists build and connect the systems, agents, and infrastructure that make the new way of working possible. People scientists bring deep expertise where the organization still needs specialization.

“I think in general… there are going to be less archetypes than we’ve had in the past.”

The bigger message isn’t that every HR team needs to reorganize into these three buckets tomorrow.

It’s that AI is already eroding many of the handoffs and boundaries that created traditional HR silos in the first place.

‍