Tags: analytics dashboard, Power BI
One of our customers, a major software provider in construction, agriculture, and geospatial industries, has one of the leading solutions for building design and building information modeling (BIM). Just think about the complexity of designing a building and the complexities of designing, fabricating, and installing air ducts, plumbing, electrical wiring, etc. In order to improve the software for their customers, their software product that we’ll call “BIM-APP” tracks all the user interactions and any system errors to log files. The development team for BIM-APP can then use the data from the log files to prioritize bug fixes and feature requests based on the number of occurrences different user interaction events. As part of their annual Hackathon, DB Best teamed up with Microsoft and the development manager for BIM-APP to a Microsoft Power BI dashboard to aid the team in designing new features that will continue to “wow” their customers.
Original customer’s system
The BIM-APP has hundreds of users all over the world using the app every day. This app generates an active log stream with valuable analytics data. The customer uses the Sumo Logic platform for data analysis and creating dashboards. Sumo Logic is a cloud log management and metrics monitoring solution. Our customers appreciated the capabilities of Sumo Logic and the ease of processing their log files. Unfortunately, this solution limits the data history to just 30 days.
The main drawback of the existing solution is that the charts can only go back 30 days. That’s why our client wants a better look at their analytics data. They considered creating an alternative reporting platform based on Tableau or Microsoft Power BI. The customer wanted to take advantage of the previously unused data, looking for a solution that would allow for analyzing months or years worth of data.
Creating a Power BI dashboard
As a part of Proof-of-Concept, we created a comparable Microsoft Power BI dashboard. The Power BI solution consisted of a single Power BI report (PBIX). The PBIX doc contained all the queries, datasets, functions and parameters needed to parse and display the data. Learn more about how we approached the task.
1. Data cleanup
First of all, the customer provided us with a Sumo Logic log file in a CSV format. This means that at the end of the day we can easily compare the customer’s existing Sumo Logic dashboard with the Power BI solution crafted by DB Best.
This source file contained various data elements. We needed to parse the valuable data from the JSON for analysis. We leveraged the Power BI interface to parse the data into manageable columns. The customer was happy with the ease of using Power BI for data cleanup.
2. A deeper look at the data
Microsoft Power BI provides users with many additional ways of looking at the data. The interactive charts are updated automatically once you apply the filter by clicking on the diagram elements. The interface proved to be very easy. Basically, all the customer needs to do is simply drag elements from the data onto the visualization board. Then you can simply watch how Power BI automatically cuts up your data into various charts, graphs, and many other visualizations.
Also, we created an animated chart, allowing to go along the timeline as the data moves around to display the features being used to what extent over time. The customer enjoyed the visuals, that we created, in particular, and the design process in general.
3. Using geolocation data
Looking at the raw data provided by the customer, we discovered that it includes users’ IP addresses. Our customer never cared about this data, treating the user’s IP addresses as the “dark data” in their Sumo Logic dashboards. We showed the customer how they can make this data work and brought it back to life by using it in Geolocation web service API.
We created a custom function in Power BI, which allows to easily source the IP addresses of various data points within the client’s API. Using this data, we could break down exactly where in the world the data was coming from. With this newly formed data, we created visually appealing maps which segmented data into different countries. The client will save a substantial amount of money by being able to specifically target certain areas or customers. Using this data, they will no longer needlessly waste money on broad ad campaigns.
The customer liked this approach very much because it allows optimizing targeting customers throughout the world.
Sumo Logic dashboard vs Power BI dashboard
By the end of the day, we clearly convinced our customer that Microsoft Power BI was more capable of analytics than Sumo Logic. Power BI provides the customer with the following advantages, which turned out to be crucial.
- Auto insights — Power BI looks over your data, uses Cortana AI to give you several dashboards of different insights and various correlations of your data that you may not have thought of.
- More dashboards with animations, automatic filtering, easy interface, and many more ways to view your data in interesting ways.
- The ability to use SQL Server, MongoDB, and pretty much any data source you can think of.
- Power BI doesn’t limit your data analytics to 30 days. In fact, it is up to you to determine the historical coverage.
- Also, Power BI provides users with an easy access to the charts via native iOS and Android applications.
And the last, but not least: Power BI turned out to be less expensive. To have a great start, the customer can even settle for a free trial period.
We managed to showcase the functionality and capabilities of Microsoft Power BI to our customer. We crafted an amazing visually appealing and easily understandable geographical dashboard. This Power BI dashboard helps the customer dive deep into their own data, gain a greater overview of their business and effectively target advertising expenses.
The Power BI dashboards, created by the DB Best team, helps visualize the dark data which was previously ignored and wasted. Now, the client is able to get a better view of their data and can get several months worth of data rather than the 30 days that they were getting with their previous solution.
In the future, we can upgrade the customer’s application to stream the log directly into the Power BI dashboard, avoiding the need to extract the data from the log file. However, we can consider storing data in the Azure SQL Data Warehouse to provide the customer with deep historical coverage.