29 Predictions: How AI Will Cement Microsoft’s Dominance and Threaten FP&A Platforms
Predictions #15–16: The FP&A Software Market Is About to Be Restructured
Disclaimer: The following views and opinions are my own and do not reflect the views or opinions of any professional organizations or enterprises I work/partner with.
Prediction #15: Microsoft Continues to Dominate the Ecosystem — and Excel Retains Its Position as the Number One Finance Tool on the Planet
There is a widespread belief in this space that AI is going to kill Excel. I want to be direct: I think that’s wrong.
I get prodded on this regularly on LinkedIn, where salespeople who work at software companies tell me that Excel can’t scale. And that for large corporations, Excel isn’t the go-to planning tool. I don’t disagree with them. Excel shouldn’t be the primary planning tool for Fortune 500s. But it is a linchpin for millions of others.
Excel isn’t going away. And the data backs this up — this 2026 FP&A Impact Report, from Vena, found that 90% of respondents are using Excel, with spreadsheets reigning supreme even inside the largest enterprises. That’s not a legacy platform dying out. That’s the current, active reality in the FP&A market. And do note that the purveyor of the report is an FP&A planning platform that generally operates independently of Excel, but likely acknowledges that Excel isn’t going anywhere.
The reason Excel survives isn’t that it’s the most technically sophisticated tool for financial planning and analysis. It survives because it’s where finance professionals already are — and because Microsoft has been methodically building AI capabilities directly into the environment rather than asking users to leave it.
I can’t share a whole lot with you about what Microsoft’s product team is building on the AI side. But I can share that they’re making the tool more powerful, more integrated, and more seamless to work with.
Microsoft Azure, Fabric, Copilot, Scout, and now integrations with Claude and ChatGPT are all moving inside Excel rather than replacing it.
The average finance professional doesn’t have to learn a new platform, with new language and interfaces. They get dramatically more powerful capabilities layered into the tool they’ve used for twenty years. 2-3 years ago, experimenters with Copilot – including me – were questioning whether Copilot was going to win the race against the Anthropics and OpenAIs of the world.
What I eventually came to accept was that the winning horse doesn’t matter nearly as much when Microsoft owns the entire racetrack.
The result is that Excel becomes far more capable and intelligent, while retaining the accessibility and familiarity that made it so dominant in the first place.
Despite my passion for Excel and embracing the joy that comes from innovating within the tool, what will change is the nature of Excel proficiency. Being technically skilled at Excel — building dynamic models, knowing advanced functions, understanding array logic — is going to carry roughly the same career weight as being a fast typist. It becomes table stakes, not a differentiator.
I say this as someone who holds a Microsoft MVP designation and has spent more than two decades teaching Excel to finance professionals. The skill doesn’t disappear in value overnight. But the premium it once commanded is going to erode. I believe the most valuable skills will shift from building models from scratch to auditing them, understanding how they work, and interpreting what the results actually mean.
The professionals who will thrive in this environment are going to be those who understand how to direct AI within Excel to do what they need and who have the judgment to know when the output is right and when it isn’t.
Prediction #16: FP&A Platforms Will Face an Existential Crisis as the Problems They Were Built to Solve Get Solved Differently
The enterprise FP&A software market has been built around legitimate problems: Excel has real limitations when it comes to dynamic, connected planning across decentralized remote teams. Rapid updates at scale. One source of truth across fragmented systems. Cloud-based architecture that needs appropriate access controls.
These platforms – many of them – solved problems that Excel genuinely couldn’t. They’ve charged accordingly, often with price tags that require significant justification from CFOs to approve. These platforms provide the engines, single sources of truth, cross-border consolidation power, and scalability that middle-market and large corporations require.
But they’re facing a new threat. After spending countless hours building workflows with Claude, I’ve become convinced that large AI models like Anthropic’s are evolving into a common intelligence layer that sits on top of existing business systems. Instead of buying another specialized application, organizations will increasingly ask whether AI can perform the same work using the data they already have. I’ve seen this firsthand by building these capabilities into my own business.
The existential threat for many FP&A platforms isn’t that they’ve stopped solving important problems. It’s that those same problems are increasingly being solved outside the platform. As AI becomes more capable, finance teams may find they can achieve many of the same outcomes using the systems they already own, rather than purchasing another layer of specialized software.
Consider the price tag most mid-market finance teams are already paying for an enterprise FP&A platform — often $30,000 to $150,000 a year once implementation and maintenance are factored in.
The quiet secret is that many of these platforms are already building on the same foundation models from Anthropic and OpenAI that anyone can access. To their credit, they’ve invested heavily in security, governance, and protecting customer data. But I increasingly believe their biggest competitive advantage isn’t going to be proprietary AI—it’s convenience. Users can stay inside a familiar FP&A environment while the platform retrieves data and invokes AI behind the scenes through MCP connectors and other integrations.
The unique functions that FP&A tools introduced years ago – multi-user collaboration, segmented administrative rights, live data connections, and audit trails – aren’t so unique anymore. When Copilot can pull month-end revenue and reconcile it against a forecast inside Excel or Dynamics, the CFO who was looking at standalone FP&A tools might second-guess the decision.
These products face a fundamental challenge. They can either become a legacy system for existing customers with limited growth prospects, or they can reinvent themselves as AI-enabled systems and find new customers and new use cases.
Don’t believe me? Google some of the leading FP&A tools in the market and you’ll see that their taglines have changed. They used to be FP&A tools – now they’re AI-native, or AI-first, or full end-to-end operating systems. They’re repositioning and rebranding.
Today, the narrow moat many of these platforms have isn’t superior technology—it’s switching costs. Implementations are expensive and data migration is painful. Finance teams have spent years building processes and reports around these systems. That’s enough to make many CFOs postpone replacing them.
But switching costs aren’t a durable competitive advantage. They delay change; they don’t stop it. As AI-native alternatives continue to mature, the justification for the high price tags becomes harder to make, especially for mid-market companies that never needed the full complexity of an enterprise platform in the first place.
