Using E2.o in Business Intelligence or Business Intelligence in E2.o?

I recently did a Google search to see if I could find anyone who has written on Business Intelligence and Enterprise 2.0. I came across a series of articles written by Colin White on the b-eye Network on just such a topic. I commend Mr. White in doing the difficult job associated with building out a framework for E2.0 and BI. I now have the advantage of commenting and building upon what he has written. Hopefully my comments in this post will not be taken in anyway disparaging what Mr. White has written. I am responding with hopefully semi-coherent thoughts, since the article created some level of emotion that I felt compelled to make this post. As I said, I have the easy job of writing the blog post!

The article starts out stating “This series of articles examines the use of Enterprise 2.0 in business intelligence (BI)”. This may be semantics, but I see the need to make the statement in reverse, “the use of Business Intelligence in Enterprise 2.0”. When I envision an Enterprise 2.0 platform, I see business intelligence as a plug-in to the E2.0 platform. It is analogous to how Facebook works. The Facebook platform provides the foundation for the social network that exists between people. Applications are then distributed and shared across the “social graph” as Facebook calls it. Relate this back to E2.0 and BI. Business Intelligence “objects”, such as a report or metric, can be distributed and shared in the same fashion. This is why I see the statement as the use of business intelligence in Enterprise 2.0.

Mr. White continues on to discuss the current challenge that BI deployments face, which is user adoption. He points out that this is driven by two primary issues: complexity of the BI tools and the ability for business users to understand the data. I agree that these two issues are #1 and 1A in why BI applications fail. So how can E2.0 address these two issues? The first is that enterprise 2.0 technologies…wikis, blogs and social networks…are inherently easier to understand and use. This is because the tools are built around driving collaboration and discussion. Having a conversation is easy, assuming the two people talking speak the same language. Enterprise 2.0 is really putting that dialog in a massively collaborative environment that is a many to many discussion. Where E2.0 can really start to drive adoption of BI is around the second issue of understanding data. By providing a platform that allows users to provide context and meaning to data, you start to address the understanding gap that exists. Simply viewing a report with data will certainly cause confusion without context. What better way to do this then by using a collaborative environment where users can provide commentary and textual descriptions.

Another concept that Enterprise 2.0 can bring to business intelligence is the “Amazon Effect”. For example, users who viewed this report also look at these other reports. Or users rated this metric a 4 out 5 when it comes to value in understanding performance, and here is why. These may be silly examples, but the concept of tapping into the collective intelligence of an organization to drive better decision making can have a powerful impact.

A challenge I foresee is how can we link the E2.0 world of unstructured data and the BI world of structured data? This is quickly becoming a challenge that is being addressed in a couple of different ways. There are niche software vendors that are providing information access and search platforms that integrate structured and unstructured data. Endeca, Attivio, Fast Search & Transfer and Northern Light, to name a few, all provide technology platforms that address this subject area. Glyn Heatley, a TalkDIG contributor, happened to provide a post on the topic of structured and unstructured data this morning. I also did a post a couple of weeks ago on the topic of building out an enterprise semantic layer through interpretative technology that can deduce the objects and their relationships based on the natural language inside of text.

On the topic of structured and unstructured data, the article discusses issues associated with accuracy, stating that “Information is never 100 percent accurate”. For business intelligence applications and deployments, success will be driven by the overall trust in the data being presented. Structured data such as revenue will need to be accurate. In an E2.0 environment, different challenges will exist. Since there is no governance in place for the unstructured data, focus will need to be on the validity associated with the context being provided.

The article concludes with Mr. White’s seven components of Enterprise 2.0 for Business Intelligence. These components include information collaboration, exploration and analysis, integration, syndication and delivery, a rich user interface, a web-oriented architecture and adopting open source solutions. Mr. White will discuss each of these components in separate articles during the series and has already completed two additional articles in the series.

Finally, the article discusses how the BI vendors are addressing the adoption of E2.0. Mr. White points out that the major BI vendors have been slow to consider and adopt the E2.0 technologies into their respective platforms. In my opinion, this is a good thing and let me explain. If you go back to how I started this post, I made the statement that I believe BI is an application service that plugs into an E2.0 platform. If you agree with that philosophy, then the BI vendors need to be looking at other enterprise 2.0 platform providers. The issue to date is that there are very little standards that would provide the plug and play of applications into an E2.0 platform, although Open Social is starting to gain some momentum. But if you take another look at what has made Facebook wildly popular and successful, it was the recent decision to open up the platform to outside developers to build applications to run inside Facebook. This is what is needed for E2.0 to elevate within the enterprise and start to see the user adoption expectations associated with business intelligence.

So what are your thoughts on the topic? I have in no way answered all the questions. If anything, I have probably only left more questions to be answered. Hopefully that is the case and we can have a dialog on the topic.