Azure DataMarket: a postmortem and a proposal

Microsoft are retiring Azure DataMarket due to a lack of sustained customer interest in DataMarket. Azure DataMarket was an online store of data, which users could rent or buy data. Why did it fail, and where do we go from here? This is a shorter version of my longer blog here.

Master Data Management, interrupted

I hoped the Azure DataMarket might become a cloud-based Master Data Solution cog in the enterprise wheel for business intelligence and business analytics projects. I'd seen it's value in offering 'single version of the truth' data. In my mind's eye, business users could download 'packs' from the Azure DataMarket which could form the beginnings of a Geography dimensions, Data dimensions and so on. This would streamline data across the whole organisation as part of an enterprise process, and would potentially help data to be burst out of the silos where it's buried. Organisations don’t seem to consider Master Data Management very much, and it's hard to articulate its value and ROI when organisations simply muddle along with vertical LOB-based data silos.

A solution looking for a problem

Having used it for some time, Azure DataMarket seemed to be aimed primarily at IT people, not business people. IT people are pretty good at searching but business people don't always understand wildcards and so on. Azure DataMarket connected to Power Pivot and Visual Studio; it was aimed at the wrong audience.

The problem Microsoft should have been trying to solve

Help business people search for data and information, and have 'packs' of data in Azure, similar to content packs for Power BI. This would increase uptake of Azure amongst business users, who are a nebulous bunch at the best of times.

Let's hope we see Azure DataMarket back for the right audience, not for the people who didn't really need it.