Frontier Firms: The Strategic Shift to Agentic Business Applications
Bryan Goode has seen plenty of technology transformations during his 21 years at Microsoft. The shift from on-premises to cloud computing. The evolution from single-player to multiplayer productivity tools. But this year feels different.
“This one is a whole different thing,” says Goode, Corporate Vice President at Microsoft. “There’s an opportunity for customers and companies to reinvent their business with AI at the core.”
The difference isn’t just about adopting new technology. It’s about building a fundamentally different type of company.
Why This Time Is Different
Previous transformations changed how businesses operated. Cloud computing moved infrastructure. Collaboration tools changed workflows. But these shifts didn’t fundamentally alter what companies were or how they worked.
Agentic AI does.
“Those other changes were transformational in many ways, but they didn’t really give rise to different types of companies that worked in different kinds of ways,” Goode explains. “But I think AI and agents are giving rise to that. It’s not just a change in the tech. It’s actually a change in the company and the customer supported by the tech.”
That’s what makes frontier firms worth watching.
What Defines a Frontier Firm
Frontier firms aren’t just using AI tools. They’re integrating AI into their core operations and business models.
“I look at frontier firms as a bit of a living lab,” Goode says. “These are companies that have put AI within the core of their operations and their business model.”
The distinction matters. Any company can start using AI. That’s a good first step. But frontier firms are operating at a different level entirely.
They’re the ones defining what’s possible. And there’s a lot to learn from how they deploy AI, what patterns they use, where they see success, and where they struggle.
Three Patterns of Work
Goode identifies three distinct patterns emerging from how organizations use copilots and agents:
- Personal productivity is the entry point for AI-powered innovation. This is where individuals use AI to work more efficiently. For example, a salesperson uses Copilot to summarize meeting notes alongside a support agent that helps draft customer responses.
- Process improvement takes innovation further. Here, agents automate specific business processes. An agent handles routine customer service inquiries. Another manages invoice processing in finance. Both escalate to humans when clarity is needed, or processes require a person to step in.
- Functional transformation is where things get even more interesting. This is when entire business functions are redesigned around AI capabilities. The third phase of a Frontier Firm is when humans set direction while AI agents run entire business processes and workflows, checking in only as needed.
“We’re starting to see customers who are saying, ‘I’m going to completely reimagine my sales function, or I’m going to completely reimagine customer service,'” Goode explains. “They’re not just adding AI on top. They’re fundamentally rethinking how work gets done.”
From Systems of Record to Systems of Action
For decades, businesses have run on systems of record. These databases store financial information, customer data, and operational details. Workflows wrap around them. People update records, but they didn’t really work in these systems.
And these systems are often siloed. Sales data sits separate from financial data. Customer information lives apart from operational metrics.
“That’s really the way business has worked for decades,” Goode says. “But that’s been pretty static. People mostly don’t want to go into those systems of record to do work. That’s not where work happens for or by them. That’s actually more where reporting happenss.”
The shift to systems of action changes this, fundamentally.
In this new model, work actually happens in these systems. Data integrates across business functions. Structured data from line-of-business apps combines with unstructured data from productivity tools. Agents operate within this integrated environment.
“You now have something totally different from the way we’ve run businesses for the last five decades,” Goode notes.
This architectural shift isn’t just about technology. It’s about enabling entirely new ways of working.
What Functional Transformation Looks Like
Microsoft has been running its own transformation as a test case. The company deployed agents across various functions and measured results carefully.
In every case, they saw immediate improvement. But the real gains came through iteration.
“We started, we picked a measure, we watched the measure, we saw improvement immediately,” Goode says. “But the real results came through iteration. Now that we’ve got an agent doing this, what if we had a human do that? Or what if we added another agent to complement that first agent?”
This iterative approach is what separates organizations seeing AI as productivity enhancement from those redesigning entire business functions.
Customer engagement appears to be moving fastest. Almost every one of our customers has started some kind of customer experience transformation project using AI. The reason is simple: you can see immediate impact. Take costs out of support. Help sales teams grow revenue.
But Goode is clear, this is just the beginning. Every industry, every company, every function is being transformed right now.
The Gap That Matters
There’s a crucial point that often gets missed in discussions about AI adoption. As Goode puts it, businesses aren’t failing because they can’t use AI. They’re failing because they can’t integrate their siloed data.
Without integrated data, AI can’t deliver its full value. That’s the foundational work that needs to happen first.
For organizations trying to become frontier firms, the data integration challenge is real. But so is the opportunity.
The companies getting this right aren’t just adopting tools. They’re building fundamentally different organizations.
That’s what makes this moment different from every other technology transformation.
For more information please visit https://www.microsoft.com/en-us/power-platform