Here at BizData, Power BI is frequently an important part of the solutions we deliver to our clients. Because of this, we keep a close eye on updates and developments on the Power BI front. It has been an interesting year so far for Power BI feature updates and in this post we’ll go over exactly why we’re excited about some of the announcements and how they are relevant to our customers.
For large organisations, managing large numbers of users in workspaces has been made significantly easier. With this crop of change, workspaces have been untethered from Office 365 Groups, opening up a range of new sharing and access options to workspace admins using existing processes such as security groups or distribution lists. The list of users who get notified about workspace changes can now also be customised and a new UI allows managers to view useful usage metrics on any workspace. Additionally, OneDrive access from workspaces has been made more customisable, allowing for more flexibility in managing team data storage.
Power BI is now previewing an unprecedented access to Azure Machine Learning capabilities for the "Citizen Data Scientist". With the new system called AutoML, a broader range of users can pick up and apply machine learning to datasets, extracting insights that had only been available via separate Azure services. The automation and detailed user interface allow for the training and refinement of a ML model directly within Power BI. With such a powerful tool being made more accessible, a broader range of analysts will be able to pick up an apply these solutions to a broader range of data.
Dataflows are a feature of Power BI that focuses on providing analysts with more powerful self-service data preparation capabilities than is available on the desktop. While this is not an area of interest for the average business user, there are a number of changes in this area of Power BI that are worthy of note. Firstly, there is now the option for alert emails to be automatically sent when there is a data refresh failure. This means more transparency when things don’t go to plan and faster turnaround times for issue resolution.
Additionally, there is a new option for easier dataflow handover, meaning that ownership of dataflow management can be passed on easily when required. For projects that involve multiple stages or evolve into new forms, with new consultants taking over, this means a decrease in overall friction when handover is involved.
Finally, the option to edit workflow container size gives workspaces that are dealing with very large and complex dataflows better resource management in order to potentially improve overall performance. For anyone trying to optimise the performance of data refresh this is a very welcome feature addition.
The general availability of Multi Geo is a change which mostly takes place behind the scenes, but it does have the potential to positively affect all Power BI users. By being able to optimise where, geographically, cloud data is stored, there can be a noticeable improvement in refresh times of datasets.
Another advantage of this change relates to organisations who have limitations on where their data has to be stored geographically. With the Multi Geo feature, they can now use Power BI and be in full compliance with relevant data security restrictions.
Typically, when given the choice between generic and less-generic, the latter is clearly the way to go. With newly introduced options for tailored links, we now have the option to replace the generic Microsoft help portal and support links in Power BI with options that are more directly useful to a particular organisation. By leading Power BI users to a specialised organisational resource centre, we can deliver a repository of useful material that can be used as an easily accessible reference directly from the Power BI interface.
The ability to export a power BI report as a PDF or a PowerPoint file is now finally implemented for all users and the page limit has been increased to 30. With the relative popularity of using PDFs in report distribution within organisations, we anticipate that this will please those who would rather not load up Power BI every single time they want to view a certain report.
Another welcome change is the improvement to the email subscription feature. This addition enables users to send out a batch of emails on demand, at the click of a button and notify all report subscribers, bypassing the usual automation. This is useful for the purposes of testing or flagging particular subscriptions.
A new standalone Windows application, the Power BI Report Builder (familiar to those who have traditionally used SQL Server Reporting Services) allows for the creation of paginated reports that can be published on the Power BI service and easily exported to a variety of formats. In addition to producing reports that are visually appealing and “pixel-perfect”, they can be optimised for printing. This is great news for anyone whose operational requirements require print-outs or offline reports. While not everyone’s cup of tea, providing the choice between interactive and static reports is bound to cover more reporting scenarios for organisations.
Another change that will be exciting for those who appreciate a solid, well-put-together report export. With filtered export capabilities, any report export now has the ability to carry over all the slight adjustments you currently have active in the current viewing of the report. A new dropdown allows you to carry over into your export the currently applied filter, slicers, sorting, drilldown, focus and visibility for the current report. With this change, it is easy to generate a report that is zeroed in precisely on the elements that you want to focus on.
The Key Influencers Visual is the first “AI Visual” providing machine learning powered feedback directly within the report visual design canvas.
It automatically produces a visualisation that breaks down the relative importance of each data variable on a particular outcome (this may be a status, category or a numerical value), so that you can easily get a birds-eye view of many variables at once (instead of being limited to choosing two or three that can fit on a traditional chart). Behind the scenes, it automatically runs a logistic or linear regression and decision tree to make it easy to quickly understand factors that may warrant further investigation.
If you want to become more acquainted with these recent updates or simply want to brush up on your Power BI fundamentals, BizData offers training sessions that can help you do both. Click here to find out more.