I recently read an article by David Strom titled “The Best Self-Service Business Intelligence Tools of 2015.” It made me think about the role, purpose, and practical use of self-service business intelligence (BI) tools for decision support in organizations. There is a lot of hype and confusion around self-service BI and what some industry analysts call “traditional BI.” [Disclaimer: I am vice president of Product Marketing for Information Builders, an enterprise business intelligence software company.]
What drew my attention in the article was Strom’s statement that: “When most people first think of self-service BI tools, they think of using a spreadsheet for their data analysis and graphing needs.” This is absolutely correct and indeed Excel has the largest use and share for individual data analysis. Then the author proceeds to explain how the new self-service tools provide enhanced capabilities compared to Excel to work with larger volumes of data, to combine multiple data sources and to visually explore data for patterns in an easier way. The benefit for front line managers is that the tools are easier than traditional BI tools, so if they have questions about the business then it is easier to use this DIY approach to get answers.
But is the DIY approach the right approach for front line workers to get answers to business questions? Excel has been around for years and its popularity is not diminishing. The drawbacks of Excel have also been known for years, as have the drawbacks of traditional BI. Is making a better DIY tool going to attract more users and drive better decision making?
To answer these questions we have to approach the problem from a decision making perspective. That means we have to understand not only what information front-line employees need but also how it needs to be presented to them in order to empower more fact based decision making. Decision theory and behavioral economics are two branches in science that study how people make decisions in order to improve the process and eliminate biases. We know from these studies that the human brain has two systems for making decisions. The Nobel Prize winner Kahneman calls this System 1 and 2 Thinking. System 1 is fast and automatic and supports immediate real time decisions. System 2 is slow and deliberate and helps us take into account all intricacies for strategic long term decisions. Hence, there is a fine balance between the two systems. While system 1 avoids analysis paralysis, system two prevents some errors.
Let us look now at organizational decision making. Front-line employees, also referred to as operational employees, have to make fast decisions to capture opportunities or eliminate errors. This is why front line decision making is largely standardized and routine. The front-line operations of organizations rely on System 1 to make decisions. You can see the validity of this by asking retail executives what they expect of store managers – to stay in the back office and analyze data or to be on the shop floor managing the operations and interacting with customers. Everyone wants their store managers to be on the shop floor. Hence, there is a tension between how much time people can spend analyzing data vs. how they can get answers to business questions quickly in order to make decisions instantly. Toyota has captured this sentiment and put in the Toyota principles that no information system should cause the engineers to go and ‘play’ with data to get insights. The engineers should receive clear information status in order to be able to make decisions instantaneously. Or in other words, front line employees would benefit most from BI if they get answers to questions instantly, and NOT if they perform self-service BI. It is clear that if most employees spend 2-3 hours combining data sets and analyzing data, productivity will decline by the time spent on analysis.
Hence, the right approach to empowering front line employees with information for decision making is not by distracting them with self-service tools, nor by pushing more reports as traditional BI does. The latter creates information overload – a problem I will tackle in another blog. The right approach is to deliver an “app like” experience where operational users can get answers to questions in less than three minutes. In this way they can see the facts and perform the operational tasks. Productivity will increase as the time spent on answering questions decreases, but also the number of errors will decrease given the easy access to facts.
Thus, any approach has to be judged against how people make decisions and how the decision making impacts the productivity of the organization.
This article is published as part of the IDG Contributor Network. Want to Join?