Microsoft just dropped its fifth annual Work Trend Index report called “2025: The Year the Frontier Firm Is Born.” This 15 page behemoth is packed with insights from 31K workers across 31 countries.
The big takeaway? We're witnessing the birth of a completely new kind of company, one that's powered by “intelligence on tap” and hybrid teams made up of both humans and AI agents. Whoa, dude.
We went through the full 15-page report, and even got to chat with Microsoft's Chief Marketing Officer, AI at Work Jared Spataro, and distilled the most important points that matter to you.
The Frontier Firm: What Is It?
Think of Frontier Firms as companies that are all-in on AI—not just experimenting, but integrating it throughout their entire structure. They're defined by:
- Organization-wide AI deployment.
- Advanced AI maturity.
- Current agent use.
- Plans for extensive agent integration.
- A belief that agents are key to ROI.
The surprising part? Out of 31K survey respondents, only 844 worked at companies that meet this bar. But these early adopters show where everyone else is heading.
Workers at these Frontier Firms report some pretty striking differences:
- 71% say their company is thriving (vs. just 37% globally).
- 55% can take on more work (vs. 20% globally).
- 90% say they have opportunities for meaningful work (vs. 73% globally).
- 93% are optimistic about future work opportunities (vs. 77% globally).
Three Big Shifts That Are Happening Right Now
1. You Can Buy Intelligence On Tap
The report introduces this concept of “intelligence on tap”—the idea that intelligence is becoming abundant, affordable, and available on demand.
Here's a unique problem Microsoft's research exposed: There's a massive gap between what businesses demand and what humans can sustainably deliver. The stats are pretty wild:
- 53% of leaders say productivity must increase.
- 80% of the global workforce (both employees and leaders) say they lack enough time or energy to do their work.
- During the 9-5 workday, employees are interrupted every 2 minutes by meetings, emails, or pings.
- PowerPoint edits spike 122% in the final 10 minutes before a meeting (we've all been there).
- 60% of meetings are ad hoc versus scheduled.
- Chats outside the 9-to-5 workday are up 15% year over year.
The solution? AI agents that can reason, plan, and act as digital labor, allowing companies to scale capacity as needed. 82% of leaders are confident they'll use digital labor to expand workforce capacity in the next 12-18 months.
2. Human-Agent Teams Are Upending the Org Chart
Microsoft suggests that the traditional org chart is being replaced by a “Work Chart”—a dynamic, outcome-driven model where teams form around goals, not functions.
In this new model, expertise is on demand through AI, so companies can spin up lean, high-impact teams as needed. Nearly half of leaders (46%) say their companies are using agents to fully automate workflows or processes.
When asked why employees turned to AI instead of a colleague, the top reasons were:
- 24/7 availability (42%).
- Machine speed and quality (30%).
- Unlimited ideas on demand (28%).
What's interesting is that the least common reasons involved avoiding human traits like impatience or judgment—suggesting people prefer using AI not to replace human value, but to enhance it.
In our convo with Microsoft's Jared Spataro, he explained the concept of the “human-agent ratio” – a new metric companies need to consider:
“We encourage our customers to take a step back and say, well, time out, you know, let's take a look at how many people you have at your units of production, what you produce per head, essentially. And let's think about what you could do if you could augment that with agents,” Spataro said. “Could you increase those units that you kick out by adding capacity through agents? And that's pretty easy to do.”
3. Every Employee Becomes an “Agent Boss”
Here's where it gets really interesting. The report introduces the concept of the “agent boss”: someone who builds, delegates to, and manages agents to amplify their impact.
The surprising part? There's a significant gap between leaders and employees in this area:
- 67% of leaders are familiar with agents vs. just 40% of employees.
- 79% of leaders believe AI will accelerate their careers vs. 67% of employees.
According to Spataro, working with these AI tools is creating a shift in skills toward generalist capabilities: Traditionally, delegation meant assigning tasks to people. This shift requires us to rethink how work gets distributed.
With AI agents, delegation becomes more about:
- Task Decomposition: As Spataro puts it, you need to “cut up your work” into components that can be effectively handled by AI systems.
- Quality Control: As a human, YOU still need to “pass judgement” on the output, determining what meets standards and what needs revision.
- Integration: Finally, you “bring that together and produce it as a whole”—combining the AI-generated components into a coherent final product.
A New Management Paradigm
This approach to working with AI tools creates what Spataro describes as leadership advantages. He notes that “we're seeing leaders perform better on all the various dimensions” when they master this new form of delegation.
Why? Because “using these agents is like delegating and managing a team”—but with different dynamics. An AI team requires different management techniques than a human team, creating opportunities for those who adapt quickly.
Career Implications
Particularly interesting is Spataro's observation about who's adopting these skills fastest. He specifically mentions professionals “mid or late in career willing to engage having the skills.” (sounds like a good portion of our readership! If that’s you, give us a shout out). This suggests that experienced professionals who embrace AI delegation may have significant advantages.
How Leading Companies Are Using AI Agents
Here’s what you really want to know: who are these companies using AI, and what are they ACTUALLY doing with it?
Several companies are highlighted in the report for their innovative use of AI:
- Wells Fargo: Built an agent for 35K bankers across 4K branches. 75% of searches now happen through the agent, cutting response times from 10 minutes to just 30 seconds.
- Dow: Deployed agents to find hidden losses in shipping operations. Expected to save millions in the first year. Spataro mentioned that Dow has created “two custom agents with their Copilot.”
- Bayer: Researchers in Crop Science R&D each save up to 6 hours per week using an agent.
- The Estée Lauder Companies: Created an agent to identify and consolidate consumer insights.
- Holland America Line: Implemented an agent concierge that instantly responds to cruise guests.
The Secret to Getting the Most from AI Agents
One of the most practical insights from our chat with Jared Spataro is about grounding AI on high-quality input:
“One of the little-known secrets for how you get an agent to do really interesting things is you just ground it very specifically on really high-quality input content,” he explained.
For marketers, this means starting with the best brief possible—as if you were writing for a high-quality agency—then asking an agent to perform specific tasks using that brief as input.
“It's amazing... That's typically what blows people away is how good the tech has gotten. If it is well grounded on something that is very accurate, a good brief for a new product coming out can produce incredible results,” Spataro said.
The formula, as he puts it: “My people should be producing small numbers of really high-quality content. And then the machines, in this case the agents, they can go do a bunch of the kind of scale-out work that we typically have had to add bodies to do.”
We’ve found this to be true in our work, too: garbage in is garbage out, but the reverse is true, too: GOLD in is gold out.
For example: think of the brief Jared described above as your prompt. We often think of prompts as this thing we’re creating just for AI to use, but when you think of a prompt like a creative brief, it makes the prompt feel more human and less machine.
So what makes a good brief?
Just like a creative brief guides an agency, a well-crafted prompt steers an AI. Here's how traditional brief components translate to effective prompting:
- Background: A good brief provides context about the specific task, not your entire company history. Tell the AI why you need this output and what problem it's solving. The more focused your background, the more targeted the response.
- Background → Context: As MIT Sloan notes, “provide context” is one of the three key strategies for prompt engineering. Tell the AI exactly what problem you're trying to solve.
- Target Audience: No creative brief is complete without understanding the target audience. So be specific about who will consume the content. Instead of “general readers,” specify “first-time home buyers with tech backgrounds who need jargon-free mortgage explanations.” This specificity helps the AI adopt the right tone and vocabulary.
- Target Audience → “Assigning a role”: Assigning roles to AI is one of the effective techniques for prompt engineering, although we’ve seen arguments for and against roles (some say they’re necessary, others don’t). But when you specify who will consume the content you’re creating, it helps the AI adopt the appropriate tone and vocabulary.
- Business Objective: This helps define your success metrics clearly. What should this content achieve? Generate leads? Educate users? Save time? The AI needs to understand your goals to optimize its response accordingly.
- Business Objective → Clear Deliverable: To get the best results out of AI, you need to define clear parameters for the task at hand. By telling the AI what success looks like with measurable outcomes, you get yourself much closer to the end product you want with minimal edits.
- Insight/Truth: Share the key insight that should drive the content. This is your “a-ha” moment or “big idea” (as we call it in copy land) that gives the AI a unique angle. “Our customers don't lack motivation to save; they lack clear systems” is more valuable than “people want to save money.”
- Insight/Truth → Knowledge Activation: Providing the AI with enough background information to understand the scenario helps achieve specificity in the output, which is vital if you want to steer clear of generic, “AI sounding” results.
- Main Message: Distill your core message into a single, concise statement (ideally under 8 words). This becomes the North Star for the AI's response, ensuring everything supports this key takeaway.
- Main Message → Output Definition: A key 2025 best practice for prompting is to “define the desired output” by specifying “the format, structure, and style of the output for precise results.” This focuses the AI on your core message.
- Mandatories: Specify any requirements or limitations: word count, style guide rules, required terminology, legal disclaimers, or formatting needs. This prevents rounds of revision and ensures the output fits your parameters.
- Mandatories → Parameters & Constraints: The better you define and specify your expected constraints like word count, style guides, or a desired output format, the more likely the AI will produce an output in line with what you’re looking for (specifying word counts don’t work 1 to 1, but it helps get the output into the right ballpark range).
The point? Focus all your human creativity, judgment, and reasoning on creating the best possible brief (a.k.a prompt)—which the AI can then translate into all the different content types you need.
Stop repeating yourself and creating multiple versions of the same type of material. Use the AI to help you translate
What Does All This Mean For YOUR Career?
Let’s be honest—we’re all scared about what all this AI stuff means for our jobs, and whether or not we’ll still be employable next month let alone ten, twenty, or thirty years from now.
If you want to stay ahead of this “frontier firm” shift, the report suggests several actions:
- Develop an agent boss mindset: Start thinking about how you can delegate to and manage AI agents to amplify your impact.
- Focus on AI literacy: According to LinkedIn, AI literacy is now the most in-demand skill of 2025 (more on this below).
- Cultivate uniquely human strengths: Skills rising in demand include conflict mitigation, adaptability, process automation, and innovative thinking.
- Rethink your workflows: The biggest gains will come from reimagining how work gets done, not just using AI for existing processes.
- Get hands-on experience: As Spataro told me in our conversation:
“Do everything you can to get real world in-business experience with the technology. Talking to hiring managers, they just want people with practical experience. People who have practical experience, as much as you can do to integrate it into your work, you'll be on the cutting edge in a week if you're smart and serious.”
That’s solid gold advice right there—Microsoft’s report included some other helpful tips, too.
P.S: Alongside the research, Microsoft also announced new Copilot for Microsoft 365 capabilities that'll be coming up in May that you can read more about here.
Spataro emphasized the importance of Copilot Studio, calling it “the Excel of the AI era", a tool that “puts agent creation within the reach of a regular business user. If you can use Excel, you can totally use Copilot.”
And if you don't want to limit yourself to only Copilot Studio, this is a great roundup of other tools (thanks, Perplexity!) to check out as well.
Our Take
The message is clear: 2025 is going to be a pivotal year for AI in business. Companies that are all-in on AI are already seeing real benefits—higher engagement, more capacity, better work.
But there's a clear gap between leaders and employees in understanding and embracing these tools. Organizations that invest in AI skilling and bridge this gap will likely come out ahead.
The report's message feels a bit like standing at the cusp of the internet age: Imagine knowing what you know today just before the internet changed everything. That's where we are with AI—and having that knowledge now gives you the power to lead rather than follow.
So if it takes about a week’s worth of serious practice to be on the cutting edge of this industry from a hiring perspective, you’re only about a week away from being a seriously in-demand candidate in any industry. Go get your hands dirty messing around with AI, y’all!
