Power BI and the year 2019 in review

As the end of the year closes I was reminiscing on what a huge year it has been for Power BI. I work mostly with large organisations so my view will be slightly skewed towards that.

For me 2019 has been the year where Power BI got massive adoption as the standard BI platform in an organisation, it went from self serve to also contain corporate BI. These two pictures from one of Arun’s keynote presentations shows it nicely. In 2015 we were all about Data Exploration for self serve analytics.

If you now look at that picture in 2019 we have added so many elements to Power BI to make it truly a complete and Unified enterprise BI platform:

This also lines up nicely with some of the amazing features that shipped this year. Let’s take a look at what I think are some of the most important features we shipped. Yes all of these are from this year ….

• XMLA endpoint for Datasets. This feature is currently in preview. It will allow users of Power BI premium to leverage most of the functionality of Analysis Services directly in Power BI. This brings enterprise level models to Power BI and allow both IT driven models and self service models. Today you can only do Read but soon Write will be enabled too completing the story.
• Calculation groups. This is by far the most important modelling feature from the last few years. It will allow any modeler to create reusable measures and reduce the amount of measures from hundreds to tens. Marco and Albert explain it nicely too. Right now only available in AAS but with XMLA write you can do it in Power BI too.
• BYOK. This might not seem that important for most of us but when you work in the financial sector or law firms this is key. It allows you to encrypt data at rest with your own key from a key vault.
• Workspace v2. The most important Power BI feature if you ask me. This came up in every large enterprise discussion. The new workspace v2 has two major advantages over the previous version. Most importantly it doesn’t rely on Office 365 groups anymore. This means when you create a group in Power BI it doesn’t show in SharePoint or wherever else. Big IT departments argued a lot over this :). Now it just lives in Power BI by default (you can still connect them). The second one is the appearance of a “viewer role” where you can give access to someone in a group and he can only view the reports in the groups (also RLS works).
• AI features in Power BI. There have been a lot of AI features integrated into Power BI this year. It gives the power of AI into the hands of business users. There is no need for them to actually need to understand, build and maintain the actual algorithms. A few examples are the AI visual Distribution Changes analyzes what makes a distribution look different and the Decomposition Tree (my personal favorite :)) enables users to drill into any dimension to understand what is driving a key metric. With a press on the button the AI features behind Power BI can figure out what an interesting data split is. Very cool.
  The other important area here is the ability to use your organisations AI models, this has been made so it can be easily consumed by business users. Watch this space!
• Shared datasets and certification. Another great feature that allows you to create a dataset, certify it, and share it with the rest of your organisation. These shared datasets are then easily discover able. In Power BI desktop you can create the report and save it to any workspace (does not have to be the same one as the dataset lives in).
• Power BI lineage. Understanding where data comes from has always been a big ask and it is finally here. With this feature you are able to see where the data in your report comes from with a press on the button. There are also API’s available for organisations to retrieve the data for your own usage.
• Data protection. Another big one. With the increase focus on data privacy and leaks any feature that allow an organisation to manage and keep track of their data is being used, without putting it all in a safe, is incredible. The new data protection features allow organisations to label their data, create governance around it and monitor it.
• Tailoring help for Power BI users. This one one was welcomed warmly by the Power BI center of excellence at my customers. It allows organisations to add their own help sites and contact information to Power BI so they can streamline it for their organisation. Saves calls to the organisations IT team :).
• Service mail to tenant admins. Power BI now also provides incident notification so you can optionally receive emails if there’s a service disruption or degradation. While Power BI’s 99.9% service level agreement (SLA) makes these occurrences rare, it is great you now can get an email without having to get notified by your users. Another time saver for the team managing Power BI.
• Bonus: Power BI activity log. I already started this blog post when this huge feature came in. Now you can use Power BI directly to track user and admin activities. Before this feature you had to use the unified Office 365 audit log and get the right access from the Office team (which was hard). Now you can retrieve the data, as a Power BI admin, through PowerShell and store it wherever you want. Then use that to gain the insights you want and keep the data as long as you want.
• Double bonus: Azure Synapse Analytics. One of the most interesting releases of the last year has to be Azure Synapse Analytics. Of course it is still early day but the signs are very good and it gets received with enthusiasm. I also see great collaboration between the Power BI and Azure Synapse team to make the combination work smoothly. In my daily work I see more and more organisations who are moving their data to Azure. Once the data is there they want to be able to use it and gain insights out of it. That is what Azure Synapse promises. I would recommend anyone following Power BI to keep an eye out here.

Now for the new year I expect more of a continuation of this and some nice surprises as well :). It will be another great year for Power BI!