The new Microsoft Power BI
Microsoft just released the new Power BI Online: Here’s why it matters.
Jason Davenport | July 24, 2015
Today’s release of the new Power BI ushers in a new era of business intelligence capabilities at Microsoft. Power BI Online, decoupled from the SharePoint and Office release schedule, offers new data collection, cleansing, and visualization tools to business analysts that can reduce time-to-market for many reports and business-analytics requests.
Why care? Power BI’s capabilities empower business and IT analysts to rapidly develop and present analytics to stakeholders, in turn empowering/enabling businesses to understand what matters and execute an effective, data-driven strategy.
All this is possible using the current Office Excel Power Tools along with the new Power BI web platform. Using this platform and development toolset, we can reduce the report development time from weeks to days, enabling the business to build many reports with a positive return that would not otherwise be possible in a traditional development environment.
Many cloud-based tools have excellent user experiences and simplified cost and maintenance structures. These tools also have a significant ROI advantage compared to traditionally sourced and supported, on-premises tools in that they are automatically updated and do not require a company to invest in overhead to support the tool. However, reporting from these tools can be complicated—involving exports of flat files, mashing data in Excel, and performing step after manual step to create simple reports.
If enterprise BI was required, many additional technical skillsets would be required—reducing or eliminating the potential return on investment for producing a value-added report. Power BI dramatically reduces the development skills required by BI analysts. Let’s compare:
- Self-service with Power BI: Excel, Power BI Desktop
Microsoft Power BI in action: a mini case study
Power BI provides business users with the steps to automate this process entirely, using the Excel Power Tools provided with a company’s Office licenses or subscription.
For one of our clients, an apparel retailer, we needed to augment its analytics with year-over-year analysis to better predict where its decision support department should focus its resources. We built a fully-automated dashboard in Power BI that could be consumed across multiple platforms by the department’s senior leadership, with nothing more than a little Excel can-do in four steps:
- Set up Power Query to connect to the web provider via the API. We filled out a simple form on the web tool site to get a token for authentication and data collection. We then used Power Query’s simple point-and-click transformations along with a single function to query the full data source. The entire process was graphically driven and enabled us to create a repeatable, automated data collection process entirely in Excel. You can connect many on-premises solutions for automated refresh and web providers using API tokens. Power BI is also rapidly enabling new data sources by creating official connectors (or you can build your own connectors using Power BI’s REST APIs).
- Create a Power Pivot model. Luckily, the model can detect relationships from Power Query, and we only had to set up a few calculated fields to aggregate our data.
- Build Power View dashboards. These dashboards provide in-page highlighting, filtering, drill down, and cross-filtering that can be set up very quickly within the view or the data model.
- Bring it all together. We posted the Excel file to the company’s OneDrive folder, then imported the file to Power BI. Using the connection data from Power Query, we enabled automated refresh direct from the cloud provider to Power BI, and built and shared dashboards with the leadership team.
This four-stop process enabled us to reduce analyst time spent performing report creation—freeing them up for value-added analysis. We leveraged our client’s current Office 365 subscription and created a free account on Power BI. Here are some of the key savings points:
Data collection. Time saved: 2 weeks
- Prior state: Develop an ETL package to pull data using a federated account setup.
- Power BI: Pull data using Power Query from a Web API with a token. Auto refresh enabled via Power BI online.
Data modeling. Time saved: 1 week
- Prior state: Develop stored procedures to cleanse the data and create key measures.
- Power BI: Use Power Pivot to model data as requirements are gathered.
Data visualization. Time saved: 1 week
- Prior state: Develop SQL Reports to view the data.
- Power BI: Use Power View to graphically analyze the data and create visualizations.
Report sharing. Time saved: minimal
- Prior state: Post report to a company site. Users must obtain access using company accounts.
- Power BI: Post report to Power BI online. Report owner can share report with users on a need to know basis.
The next posts in this series will continue to highlight the capabilities of self-service business intelligence and explore how Microsoft Power BI can be leveraged as a rapid development and deployment tool. Stay tuned!
Jason Davenport is a Practice Area Lead in Slalom Portland’s Information Management & Analytics practice. Jason focuses on technically and organizationally enabling businesses to make data-driven decisions, utilizing lean processes and advanced analytics. Follow Jason on LinkedIn.