Ben Woodhouse, Microsoft Partner Lead, in conversation with Dufrain’s Microsoft practice. January 2026.
At the start of the year, I brought our Microsoft team together to reflect on what really matters for our clients in 2026.
A practical conversation about what we are actually seeing in delivery, in architecture decisions, and in how organisations are adopting Microsoft Data & AI at scale.
What follows is a conversation with our team, capturing their honest views on where the market is heading, where it isn’t, and what leaders should really be paying attention to this year.
Q: Ben: When you look ahead to 2026, what feels like the most important Microsoft Data & AI trend overall?
For me, the biggest shift isn’t a single technology, it’s the move from experimentation to adoption.
Over the last couple of years, many organisations have explored new platforms and capabilities. In 2026, we’re seeing far more focus on how those platforms are governed, how they fit into an existing estate, and how teams actually operate them day to day.
Fabric is still very much central to Microsoft’s strategy, but the conversations we’re having with clients are now less about features and more about operating models, cost, and long-term sustainability.
I agree with that. I think we’re moving into a much more pragmatic phase.
There will always be new announcements and new labels, but what clients are really focused on now is value. What can they realistically afford, what can they govern, and what will their teams actually use.
From a trend perspective, 2026 feels much more like a year of consolidation and adoption than a year of entirely new platforms.
Q: Ben: Fabric has been a major focus for Microsoft. How do you see Fabric really landing with clients this year?
Ian:
Fabric absolutely remains central, but the way clients are engaging with it is changing.
Many organisations are now Fabric customers simply because they renewed Power BI Premium and moved onto Fabric SKUs. What we’re seeing is that a lot of those customers haven’t yet designed what their Fabric platform should look like, how it should be governed, or how different teams should use it.
So the trend for 2026 isn’t just new Fabric features. It’s much more about platform design, adoption, data governance and getting value from what organisations already own.
Megan:
I’d add that there’s often a big gap between what’s being talked about in the community and what clients are actually asking for.
Certain features get a lot of attention online, but in our client base the questions are much more grounded. People are asking how to migrate, how to control cost, how to avoid platform sprawl, and how to stop Fabric becoming something that just “happened” without design.
That creates a big opportunity around helping clients take control of their Fabric estates properly.
Q: Ben: Microsoft announced Fabric IQ recently. How should clients think about that?
Ian:
Fabric IQ is an interesting development, but I’d describe it as something that’s still very much on the early radar rather than a mainstream adoption trend.
From what I’m seeing, there’s a lot of curiosity about what it actually means in practice and where it fits into the wider Fabric platform, particularly around knowledge graphs, metadata and how businesses find and connect information more intelligently.
At the moment, most organisations aren’t implementing Fabric IQ as a core capability. They’re exploring it, experimenting with it, and trying to understand how it might support things like copilots, agents and business discovery over time.
So, I see this less as a 2026 delivery trend, and more as a signal of where Microsoft is investing longer term in making data platforms more intelligent and more searchable for the business.
Q: Ben: There’s a lot of noise around Agentic AI at the moment. What does this really mean in practice?
Megan:
For me, the important thing about Agentic AI isn’t the label, it’s the outcome.
What we’re seeing is a growing expectation that AI should remove manual steps from everyday workflows. In BI teams, that means automating repetitive development tasks, enforcing standards through code, and accelerating delivery using tools like GitHub Copilot and VS Code.
This isn’t about taking control away from teams. In many ways it’s the opposite. The teams who adopt this well will be able to standardise how they work, improve quality, and scale much faster without increasing headcount.
Ian:
I’d agree. The successful use cases we’re seeing are very practical.
Where Agentic approaches are helping teams remove friction, reduce manual effort, and improve consistency, they land well. Where they’re positioned as something abstract or experimental, they tend not to progress beyond pilots.
In 2026, I think Agentic AI will quietly become part of how teams work, rather than something they talk about as a separate initiative.
Q: Ben: Which migrations or platform changes do you expect to matter most this year?
Megan:
There are four areas that I think will generate consistent demand.
The first is SAP to Fabric. As more organisations modernise their ERP landscapes, they’re looking closely at how Fabric fits into that.
The second is Synapse to Fabric migration. We would expect to see much more noise around that throughout the year as clients plan their modernisation roadmaps.
The third is Agentic BI, as we discussed, and the fourth is what I’d call Fabric adoption itself.
We’re seeing a lot of customers who technically have Fabric, but haven’t designed or governed it properly. They have pockets of notebooks, pipelines and workloads that nobody is really managing. Helping those organisations get back in control of their platforms will be a big theme in 2026.
Ian:
From my perspective, migrations are becoming less about moving from A to B and more about designing the future estate.
Clients are much more thoughtful now about what role each platform should play, rather than assuming there will be a single destination platform for everything.
Q: Ben: We’re seeing more clients with multiple data platforms in their estate. How do you see that evolving?
Ian:
Multi-platform estates are becoming completely normal, particularly in larger organisations.
It’s now very common to see Fabric, Databricks and sometimes Snowflake all in the same environment. The real challenge for clients isn’t choosing one platform, it’s designing how those platforms integrate and how data flows between them.
That puts a lot more emphasis on architecture, integration patterns and governance, rather than on product selection.
We’re already seeing this in live clients.
In some cases, engineering platforms are being used for large-scale processing, while Fabric is being used to enable BI teams to do last-mile transformations and analytics more quickly.
In 2026, I think the organisations that succeed will be the ones who accept that coexistence is normal and invest properly in how those platforms work together.
Megan:
Most clients aren’t trying to make a philosophical choice between platforms. They’re trying to make their existing landscape work better, with the least disruption to teams and the best commercial outcome.
Q: Ben: How is this changing the role of BI and analytics teams?
Megan:
One of the biggest shifts I see is around “Power BI as code”.
Historically, that approach was seen as quite niche. I think in certain teams it will become the default.
Using tools like GitHub Copilot and VS Code will change how BI teams develop, enforce standards and collaborate. It will accelerate delivery and improve quality, but it also requires a mindset shift.
This isn’t about automation replacing people. It’s about giving teams better tools to do what they already do, but in a more controlled and scalable way.
Ian:
I’d add that governance becomes even more important as this happens.
As platforms become more powerful and more accessible, the organisations that do well will be the ones who invest early in standards, operating models and platform design.
Final reflections: What will define success in 2026?
From this conversation, five themes stand out:
- Fabric adoption and governance will matter more than new features
- Agentic AI will succeed only where it removes real manual work
- Multi-platform estates will become the norm
- Architecture and integration skills will be in high demand
- Industry language will outperform vendor language in go-to-market success
In 2026, success will be less about chasing every new announcement, and much more about helping organisations design, govern and adopt the platforms they already have.