Is there a reason to do that, ignoring AI for a moment, to have all the data with Microsoft? What's the benefit to a customer? Why would they want to do that?

Originally, there are two main aspects in terms of the data go-to-market side of things - the analytics and the reporting. Many organizations focus initially on the data reporting, and that goes back to when they had data and data silos but weren't able to get a good picture of what was going on in their organization. Reporting was a massive issue for many companies that Microsoft work with, and their solution was promoting the benefits of them having all their data in the same data state to make it easier to get a more accurate overview of what was happening.

Is there a reason to do that, ignoring AI for a moment, to have all the data with Microsoft? What's the benefit to a customer? Why would they want to do that?

The second aspect, the AI pull-through, is on the analytics side. Without all data in one place, you can't answer forecasting questions or correlate different business activities. It's about having a historical view to understand the company and a forward-looking view with a data strategy and estate. Initially, we didn't position the end goal as the analytics piece, which was a mistake. Reporting is important but limited in achieving business outcomes. Analytics and the rest are crucial and important.