As I'm re-immersing myself in the BI market after a few years in another technology space, I've had the opportunity to meet with several industry analysts to get their perspective on the current state of the market. The overall consensus that I've been hearing is that while the market is healthy and growing, it has not yet hit its true potential or original promise. Each analyst had their own take on the symptoms - and the solutions - but a few common themes emerged.
- BI is too complex. The pendulum has swung way too far towards packing more and more features into a single monolithic application, leaving most users very confused, and leaving IT organizations scrambling trying to manage these behemoth stacks. Given the amount of BI shelfware out there, vendors need to rethink their strategies and focus more on enabling intuitive task or goal-centric BI apps vs. trying to cover all possible needs in one application. (See related post here.)
- You can't conceive all requirements at the outset. Another commonly used term for BI is decision support - meaning a system that supports decision making. This implies answering questions I don't know the answers to today. Which in turn means I'm not 100% sure where the answer lies or how I will arrive at it. (See related blog posts here and here.) By requiring a heavy metadata layer to define all possible dimensions and measures, traditional BI systems are effectively pre-determining what questions the business users can answer. This imposes significant limitations on real-world decision making where you don't know what questions you want to answer until you start looking at the data. The BI app must be flexible enough to allow business users to conceive new questions and requirements as they are using it.
- You need to start with the decision, not the data. When creating BI applications you need to be decision-centric, not data-centric. Too many BI deployments are focused on pulling together as much data as possible and shoving it into the mother of all BI applications. Instead, BI applications should be geared around specific decisions or needs and bring together just the data needed to support that area.
Here at QlikTech the mantra is "Simplifying Decisions for Everyone." We are clearly focused on the points above, making BI easier to use, more flexible, and focused on meeting business user needs. What else would you add? Where do you think we, or the market, need to go to ensure BI meets its full potential?