Microsoft has extended Power BI for Office 365 with a new update that brings forecasting capabilities to the company’s cloud-based business intelligence software.
First introduced last summer, Power BI is Microsoft’s self-service BI offering that enables users to explore and analyze data using Excel, a major pillar of the company’s popular Office productivity software suite. Users who lack specialized data analytics skills can now leverage Excel to generate reports, interactive charts and 3D geospatial visualizations.
Power BI for Office 365 finally launched on Feb. 10. Julia White, general manager of Microsoft Office product marketing, declared at that time that her company was “bringing BI to a billion users,” a reference to Office’s massive user base. Big data analytics was “not in the hands of data wonks anymore,” she added.
Now, Microsoft is providing Office 365 customers with tools that help them shift their focus to their organizations’ futures.
Available now, the new Power BI forecasting tools for the Power View component allow users to “predict their data series forward in interactive charts and reports,” announced Microsoft in a statement. “With these new Power BI capabilities, users can explore the forecasted results, adjust for seasonality and outliers, view result ranges at different confidence levels, and hindcast to view how the model would have predicted recent results.”
Forecasting in Power View leverages “built-in predictive forecasting models using exponential smoothing,” explained the company in a May 8 blog post. The capability enables Power View to “automatically detect seasonality in the data to provide forecast results from a series of data.”
The result is an interactive line chart that users can fine-tune to meet their requirement. In a video demonstration of the HTML5 version of Power View for the Excel Web App, Microsoft showed how users can simply grab a line chart’s rightmost border, which marks the timeline’s end, and extend it a year into the future. Power View then generated forecast data in real time.
An Analysis pane allows users to adjust their forecasts. Automatic seasonality detection can be overridden with the seasonality slider. Data forecasters can also set the standard deviation (0, 1, 2 or 3) and explore what-if scenarios by dragging data points up or down on the line chart (HTML5 only). What-ifs will leave the underlying data and its source affected, assured Microsoft.
Power View can also handle gaps in data. “Power View fills in missing values before forecasting,” claimed the company. “Power View can add new values up to 100 percent of the existing values, so up to half the values in the line chart could be invented.”
Power BI forecasting works only line charts that feature a single line. In addition, the “x-axis value needs to have a date/time format or be a uniformly increasing whole number” and cannot contain text or decimal numbers.