Power BI was first made generally available in July 2015. Since its launch Power BI has reached more than 200,000 organizations and to date more than 400 features have been added to the service.
But what are our top 10 Power BI features?
- User community led development: Power BI product development is driven by a 50,000 strong (and growing!) user community. The community suggest new ideas and votes on existing ideas with many of the most popular incorporated into the product within a few months (initially as a Preview feature and then as a General Availability feature).
- Complete end to end BI solution in one interface: We use Power BI Data Sets to connect to and model data, Reports to display and analyse the Data Set and Dashboards to provide a clear and concise summary of one of many Reports. The browser based Power BI Portal provides users with a place to store, manage and access all of their Power BI content as well as Excel content (including that which contains Pivot Charts, Pivot Tables, Power View worksheets and even Power Pivot data models)
- Comprehensive data connectivity options: Power BI features a large and growing list of connectors to both on-premises and cloud based data sources, including: SQL, Oracle, SAP BW, Dynamics, CSV files.
- Many data visualisation options: Power BI features many Standard data visualisations, including many you’d expect (column charts, bar charts, tables, scatter charts etc.) and some you perhaps wouldn’t (maps, funnels, gauges, tree maps etc.) This array of data visualisations is complemented by the availability of Custom visualisations developed by both Microsoft and third party developers (word clouds, waffle charts, scrollers and many many more!).
- Grouping and binning data: Traditionally report writers have been limited to the groupings incorporated into the backend solution (i.e. the database or data model/cube). But not in Power BI! Power BI allows report developers to group text values in anyway they wish, therefore effectively creating their own groupings (product/service groupings for example). The same can be achieved with numerical values, which is called Binning. Binning is useful when we want to display values as ranges and then see these ranges in our charts and tables – as opposed to seeing each individual value which can be overwhelming at times!
- Cross object filtering and highlighting (aka. cross highlighting): A Power BI Report can contain one or many Report pages (just as an Excel workbook can feature one or many worksheets) and each Report page can contain one or many Data Visualisations (charts, tables etc.). Clicking on a data point in one data visualisation filters or highlights the data in other data visualisations. Moreover, we can control which data visualisations each interacts with and how it interacts with them (i.e filtering data or highlighting data). This, particularly combined with the drill down options we have in Power BI, can greatly reduce the number of data visualisation that we need to create as users can explore and analyse data more freely and flexibly than they can in other tools.
- Slicers and Filters: A report developer can incorporate both Slicers and Filters into a report. As with the item above, the presence of slicers and filters can reduce the number of data visualisations that we need to add to a report. Filters can be applied to an entire report, an entire report page or an individual data visualisation. Users can adjust filters using the Filters Panel positioned to the right of the report. Those reporting objects that users use the most to filter data can be added to the report page itself as Slicers. Slicers commonly added to a report page include; date/time, product/service, people/team, true/false and yes/no.
- Q&A (Natural Query Language): Once we exposed data to a Power BI Dashboard (meaning we’ve published a Data Set to Power BI, created a Power BI Report over the Data Set and then pinned one or more of the Reports data visualisations to a Power BI Dashboard) we can then use the Q&A tool to query this data. To give an example, we can retrieve data from a Sales Data Set by simply typing a question such as, “show me Sales by Product Category in this Financial Year”. Power BI will then return the data in what it thinks is the most suitable form (a Column Chart, a Pie Chart, a Map etc.) and we can change and customise this. Once we’re happy we can pin our new visualisation to our Dashboard – its really that simple, very impressive!
- Mobile optimised reports: Power BI allows the report developer to build out two Report Layouts, a Desktop Layout and a Mobile Layout. Power BI defaults to the Desktop Layout when a report is viewed on a PC or on a phone/tablet held in landscape mode. The Mobile Layout is applied by Power BI when a report is viewed on a phone/tablet held in portrait mode. This allows the report developer to accommodate the needs of two types of users; the mobile user who is probably looking for headline numbers whilst on the go, and the desktop user who may want to spend more time slicing, dicing and analysing the data.
- Sharing content with Power BI and non-Power BI users: As you’d expect, we can very easily share Power BI content (Data Sets, Reports and Dashboards) with other Power BI users. These users will generally reside within our own organisation but may occasionally reside outside as well (Power BI warns us when sharing content with users outside of our organisation and this feature can be turned off completely if required). But we can also share our Power BI content with non-Power BI users. This is most easily achieved using the Publish to Web feature which allows a report to be quickly and easily added to a web page. If we want to share a Dashboard, a single Dashboard Tile (individual chart or table) or again a Report with a non-Power BI user, then we can use the REST API to embed these assets into a web page or an app.
Some honorable mentions:
- Quick insights: When a new Data Set is published to the Power BI Portal, we have the option of using the Quick Insights feature to quickly create us some charts and tables based on the data. Some of the tables and charts created by this feature will be useful and some won’t. Of those that are useful there may be some that we wouldn’t have thought to create ourselves. It’s therefore always worth trying this feature before making a final decision on what your reports should contain.
- Integration with other Office products: We would expect this, particularly with Excel, and Power BI doesn’t disappoint. Data can be quickly and easily exported from Power BI to Excel for deeper dives into larger data sets. There’s also good integration with PowerPoint which allows us to embed Power BI content into our PowerPoint slides (at the time of writing the PowerPoint feature is still in preview).
- Forecasting: if we represent data on a line chart in Power BI, with a date/time dimension on the x-axis, then we can add Forecasts to our report. We can control how far into the future we want to forecast, what level of accuracy we believe can be applied to the forecast and how many data points should be used to generate the forecast.
- Data Sets a joy to develop: I’m a BI developer, so not just a report writer. As most BI developers will agree, its the quality of the data and the data model/cube that matters the most. I’ve developed data cubes and data models in various different BI technologies over the years, and I personally find Power BI Data Sets to be one of the best, if not the best for this. Tables and columns can be renamed to whatever you like whenever you like – and don’t worry if you’ve already applied these to a stack of reports, the reports continue to display these objects with their new names. Adding new DAX based calculated columns and measures to the data set is easy thanks to the wizard that guides you through the process (similar to how it does in Excel when writing Excel formulas). Checking the data set as you develop it is easy thanks to the quick and seamless switching between the Data Set view and Report view, and data can be quickly and easily exported to Excel when investigating data anomalies.