Guided and Open-ended Analytics – Serving the Real Users of Business Intelligence


Tools and techniques for developing information strategies are being constantly devised and enhanced to help businesses and people to cope with massive levels of change. While there is value to today’s existing reporting systems and is standardized, fixed reporting, they are not only based on a static view of the business, they are architecturally rigid, utilization rates are low and rather than an effective means to cope with change they are a major impediment to it. The speed of today’s business cycles require rapid adaptation. For a reporting architecture to be useful, it must exhibit the features of both fixed, consistent reporting as well as on-the-fly improvisation. This is, as the wise man said, easier said than done.

This paper details the necessary information required to:

  • Describe the use of these tools from a cognitive point of view
  • Examine the differences between guided analytics and open-ended analytics
  • Propose best practices for deployment and use

The starting point is acquiring a deeper understanding of the relationships, causes and effects among things, and having that understanding spread throughout the organization. To do this requires, among other things, tools that provide the promise of self service to evaluate, investigate and share. However, not everyone is capable or is interested in building models or maintaining analytical applications they’ve developed.

Getting the job done, then, requires a mix of tools and approaches. For those who have an analytical perspective but not a technical one, some sort of “guided analytics” through data and models is called for. What are currently referred to as “dashboards”, represent a good combination of function and aesthetics. A smaller constituency desires to not only create analyses, but to share them with those who are not so inclined. This sort of “open-ended analytics” allows an analyst (or really anyone with the understanding of the data and relationships in the organization) to author their own analytical scenarios for themselves or for sharing with others. Ideally, systems that can serve both needs will be the winners.