The power of self-service
BI on a higher level with Visual Analytics at Belfius Insurance
Innovative IT solutions can contribute significantly to a better understanding and management of businesses. A case in point is the introduction of a self-service Business Intelligence (BI) platform at Belfius Insurance. Based on SAS Visual Analytics, the platform empowers business users to create their own reports, evaluate and correct the quality of their data and eventually discover new business insights. Project Manager Steven De Wever discusses the implementation approach and the results so far.
Business users can now filter and explore their data, drill down to every detail, surface and correct annoying abnormalities, all while discovering relationships they hadn’t even thought of before.
Steven De Wever
IT Project Manager, Belfius Insurance
Belfius Insurance is a subsidiary of the Belgian government-owned Belfius Bank and a leading force in Life and Non-Life insurance companies operating in Belgium and Luxemburg. The IT department employs 150 individuals (both internal and external resources) and manages a wide variety of business-driven IT projects. However, not all projects are born out of requests from the business. We also pursue innovation projects,” says Steven De Wever. “We closely follow what is happening on the IT market and identify opportunities within our company where we can implement some of these new and emerging IT solutions and strategies.”
Taking business intelligence a step further
An important innovation project was the development of a comprehensive self-service BI platform. De Wever explains: “Traditional business intelligence platforms only provide static views of data implemented by the IT department based on specifications from the business. A self-service or so-called agile BI platform takes Business Intelligence a step further in that it enables business users to define views and create visual reports themselves. Our end users confirm that this would be of great advantage to them.”
An intuitive interface for end users
Flexibility and user-friendliness were key aspects in the decision regarding which system to implement. “We wanted top-class visualization software that easily connects with both our operational and detailed data stores,” says De Wever. “In addition, we needed a system that provides an intuitive interface for end users to quickly create reports, charts, box plots, heat maps, bubble charts, et cetera. That is why we decided to implement SAS Visual Analytics.”
Getting rid of the spreadsheet chaos
Once the system was installed, De Wever and his team (together with consulting firm LACO) invited business users to launch ideas for analysis and reporting that could enhance their business insight and improve decision-making. “Most of them came up with some self-made spreadsheets they had been using to analyze insurance claim statistics and KPIs,” says De Wever. “These spreadsheets prove that self-service is essential for business managers: they want to be able to play around with data themselves and not be confined to IT-provided reports and views. The problem was that their spreadsheets did not zoom in on the business questions which needed answering and were difficult to adapt when reporting requirements changed. In addition, the data quality was uncertain—spreadsheets were multiplied and modified by different individuals, leading to several versions of the truth within the organization. We were quite convinced that our self-service BI platform could help solve these problems.”
A systematic approach
The crux of the solution is the clear separation of responsibilities of the IT team from those of the business. De Wever elaborates: “The IT department ensures that the correct data are in place. Then, it is essential that the business users are able to find, retrieve and display these data in whatever combination they want or need. We control the data, they control the reporting, it is essential to clearly separate the two.”
For this to happen, De Wever and his team adopted a systematic implementation approach which was followed for each set of ideas launched by the business:
- Preparation workshop—A workshop (typically half a day) is organized with business users and IT staff to precisely identify the data needed for the reporting scope at hand.
- Data preparation—IT staff makes the necessary data available in the self-service BI platform and implements automatic row-level security based on source data security levels. The business users load experimental data files and create experimental reports until they have clearly identified the data they really need. At that time, they ask IT to automate a structural load.
- Implementation workshop—A workshop (typically one day) is organized to train power users (business) in accessing the data and creating the first reports. At least one essential business report is created during the workshop.
- Create and use reports—Power users explore data and create reports, supported by IT staff whenever they need assistance. Other users are given access to these reports, including filtering, zooming in and drilling down to details. IT staff further supports the process by organizing dedicated SAS Visual Analytics user forums.
Work is more rewarding
After a few months of working this way, Belfius Insurance had already completed integrating important packages of data in SAS Visual Analytics. That included data on IT and business projects follow-up, corporate insurance sales and claims, workflow management, and process reporting. “The results are quite stunning,” declares De Wever. “Business users now do not have to wait for an IT expert to create the reports for them. This is also in our interest, since we now have more time to concentrate on strategic IT projects. Equally important, it is much more rewarding to work in this way. The user can filter and explore the data, drill down to every detail, surface and correct abnormalities, and discover relationships they hadn’t even thought of before. And they know for sure that the reporting reflects one single version of the truth.”
Create a self-service Business Intelligence environment enabling power users to create their own reports
- Empowering business managers
- Shorter delays for new views and reports
- Fast detection and correction of abnormalities
- Data preparation is a fundamental step in any data visualization project and must be controlled by IT staff.
- Integrate automatic row-level security based on source data security levels.
- Organize preparatory workshops with business users to identify the required data. In addition, organize on-the-job training sessions to teach power users how to exploit the data.