Only 22% of companies are truly winning with AI and the gap is widening fast. Here’s what Frontier Firms do differently, and how to catch up.
Almost every company you speak to right now will tell you they’re “doing AI.” They’ve got the tools, the subscriptions, maybe even a Chief AI Officer. And yet the results? Wildly different.
Here’s the number that puts it starkly: Frontier Firms (the organisations getting AI genuinely right) are achieving a 2.84x return on their AI investment. Laggards are seeing just 0.84x. That means the majority of companies are, at this moment, actively losing money on AI initiatives they thought would transform them. (Source: Lantern Studios / IDC)
This isn’t a marginal performance gap. It’s two entirely different games being played on the same field.
What Exactly Is a Frontier Firm?
The term was introduced and defined in Microsoft’s 2026 Work Trend Index, which is the most comprehensive study of AI in the workplace to date, drawing on a survey of 20,000 workers using AI across 10 countries and trillions of anonymised productivity signals from Microsoft 365.
According to Jared Spataro, Microsoft’s Chief Marketing Officer for AI at Work: “Frontier Firms are those that are deliberately designing workflows around AI to create long-term value.” (Source: Microsoft / TechArena)
The operative word there is deliberately. This isn’t about which tools you’ve bought or how large your AI budget is. It’s about how intentionally and systematically you’ve redesigned the way work actually gets done.
According to IDC research commissioned by Microsoft, 68% of organisations worldwide now use generative AI in some form, but only 22% qualify as genuine Frontier Firms, achieving returns several times greater than slower adopters. The rest are either experimenting without clear outcomes or falling further behind with every passing quarter. (Source: Microsoft Cloud Blog)
The Numbers That Should Worry Everyone Else
If you’re reading this thinking “we’re doing fine with AI,” it’s worth sitting with these figures for a moment.
According to PwC’s 2026 analysis, leading organisations report productivity gains of 25–40% in AI-augmented workflows, while the median organisation sees just 3–7%. Revenue growth from AI-enabled products shows a similarly stark divide. As PwC puts it bluntly: “The gap between leaders and laggards is not narrowing with the maturity of the technology — it’s widening.” (Source: PwC / AI Business Review)
The operational picture is just as stark. 98% of Frontier Firms are already fully operational with AI across their organisations. Compare that to just 31% of laggards and 65% of laggards who are still stuck in the planning phase, running pilots that never seem to graduate into production. (Source: Lantern Studios / IDC)
The window to catch up is not closing, but it is getting smaller!
What Frontier Firms Actually Do Differently
This is the part that matters most, because it’s genuinely actionable. The difference between a Frontier Firm and everyone else isn’t mysterious. It comes down to three consistent traits.
1. They Redesign Work — They Don’t Just Add Tools
The most common AI mistake is layering a new tool on top of an existing process and expecting transformation. Frontier Firms do the opposite: they start by asking how work should be structured if AI is a native part of it.
The results are visible in the data. Microsoft’s analysis of over 100,000 Copilot conversations found that 49% of all AI interactions support genuine cognitive work: analysing information, solving problems, evaluating ideas, thinking creatively. And 58% of AI users say they’re producing work they couldn’t have achieved a year ago, a figure that rises to 80% among the most advanced users inside Frontier Firms, what Microsoft calls “Frontier Professionals.” (Source: Microsoft Official Blog)
2. They Treat Governance as a Competitive Advantage
Most organisations treat AI governance as nothing more than a compliance checkbox that slows things down. Frontier Firms treat it as an accelerant.
Microsoft’s 2026 Data Security Index found that less than half (47%) of companies have fully implemented data security controls for AI. Frontier Firms build observability, Zero Trust security frameworks, and clear accountability structures from day one. The result: their teams can move faster and with more confidence, rather than constantly stopping to fix problems created by moving too quickly without guardrails. (Source: Microsoft Cloud Blog)
3. They Invest in People, Not Just Platforms
The least glamorous differentiator is often the most important. As Microsoft CVP of Employee Experience Nathalie D’Hers puts it: “Scaling AI is less about deploying tools and more about preparing people.”
Frontier Firms build cultures of continuous learning and agility, helping every level of the organisation fundamentally reimagine what their role looks like when AI is doing the routine work. This isn’t a soft benefit. It’s what converts a technology investment into a durable operational advantage. (Source: Microsoft Cloud Blog)
The Four Ways Humans and AI Are Working Together
One of the most practical frameworks to emerge from Microsoft’s 2026 research is a model of four distinct human–AI collaboration styles, each reflecting a different level of AI integration into everyday work. Understanding where you and your team sit on this spectrum is a useful starting point for any organisation.
- Author — You create content, analysis, or output with AI assistance, but the creative direction is yours.
- Editor — You review, refine, and improve AI-generated output, applying your judgement and expertise to the result.
- Director — You set goals and parameters for AI agents, then evaluate the outcomes they deliver.
- Orchestrator — You design entire systems where multiple AI agents operate in parallel across a workflow, with human oversight reserved for exceptions and escalations.
Most professionals today sit somewhere between Author and Editor. Frontier Firms are rapidly moving their key roles toward Director and Orchestrator, which is where the biggest productivity and value gains live. (Source: Microsoft Official Blog)
What Holds Most Companies Back
It would be easy to read all of this and feel behind. That’s worth acknowledging honestly, because the anxiety is real: Microsoft’s research found that 65% of AI users fear falling behind if they don’t adapt quickly. (Source: The Letter Two)
But here’s what the data also shows: the primary constraint isn’t budget, access to technology, or even technical capability. The bottleneck is workflow design and organisational change management. Companies that treat AI as a productivity add-on, rather than a fundamental redesign opportunity, consistently fail to scale results beyond their initial pilots.
The question isn’t whether your organisation has the right tools. It’s whether you have the right structure around those tools.
How to Start Moving Toward the Frontier
You don’t need to overhaul everything at once. The Frontier Firms that got there fastest didn’t start big. They started deliberate! Three practical starting points:
1. Identify your highest-value, most repeatable processes first. These are the workflows where AI redesign will deliver the fastest, most measurable returns and where you can build internal confidence before tackling more complex areas.
2. Find the Frontier Professionals already inside your organisation. Every company has a small group of people already using AI in fundamentally different ways: redesigning workflows, orchestrating agents, building repeatable practices. Find them. Learn from them. Give them a platform.
3. Measure outcomes, not activity. The companies struggling most with AI ROI are measuring usage: how many people have licences, how many prompts were sent. Frontier Firms tie AI performance to actual business outcomes: revenue impact, error rates, customer satisfaction, time-to-market.
The Bottom Line
The Frontier Firm isn’t an elite category reserved for Big Tech. It’s a direction, which is available to any organisation willing to think seriously about how AI changes work, not just whether to adopt it.
The companies pulling ahead right now aren’t smarter or better-resourced. They started designing deliberately, a bit earlier. The question isn’t whether your organisation will eventually use AI seriously. It’s whether you’ll still be able to close the gap when you finally do.
Because if PwC and Microsoft’s data are right, the gap isn’t waiting for you to catch up!
Related article: Agentic AI: The Future of B2B Productivity
Frequently Asked Questions
Q1: What is a Frontier Firm?
A Frontier Firm is a company that has moved beyond AI experimentation to systematically redesign its workflows, operations, and culture around artificial intelligence, generating measurable, organisation-wide returns as a result. The term was defined and popularised by Microsoft’s 2026 Work Trend Index.
Q2: How many companies are actually Frontier Firms?
According to IDC research cited in Microsoft’s 2026 AI Decision Brief, only 22% of organisations worldwide currently qualify as Frontier Firms. The majority are either in a neutral experimentation phase or actively lagging behind.
Q3: What ROI do Frontier Firms achieve compared to laggards?
Frontier Firms are achieving an average 2.84x return on AI investment. Laggard organisations are seeing just 0.84x, meaning most of them are, on net, losing money on their current AI programmes.
Q4: Is becoming a Frontier Firm only possible for large enterprises?
No. While larger enterprises have more resources to deploy at scale, the defining traits of Frontier Firms (deliberate workflow redesign, strong governance, and a culture of continuous learning) are accessible to organisations of any size. The key is intention, not budget.
Q5: Where should a company start its journey toward becoming a Frontier Firm?
Start by identifying your highest-value, most repeatable internal processes. Look for the employees already using AI in sophisticated ways and learn from them. And shift your success metrics from AI usage to actual business outcomes.
Q6: What are the four human–AI collaboration models?
According to Microsoft’s 2026 Work Trend Index, the four models are: Author (creating with AI), Editor (refining AI output), Director (setting goals for AI agents), and Orchestrator (designing multi-agent systems). Most organisations are currently at the Author and Editor stages, while Frontier Firms are advancing toward Director and Orchestrator.