Is It Time To Rethink the Definition of Business Intelligence?

A few weeks ago I, along with my sometime partner in crime, John Crupi, had the opportunity to sit down with some of the nice people at the Boulder BI Brain Trust (BBBT). There were some very big Business Intelligence thinkers in the room, including Claudia Imhoff and Barry Devlin. Walking into a room of big-wigs like this is never easy, but in this particular case I was more apprehensive than usual because we were going to discuss our interpretation (reinterpretation, perhaps) of the fundamentals of BI. And I wasn't entirely sure anyone would agree with us.

If you want the highlights from the discussion, check out John’s podcast interview with Claudia. For my purposes today, I'm going to cut straight to the end: I was wrong to be apprehensive. We went in thinking they’d call us BI heretics but instead we were greeted with open minds and enthusiasm – it was a healthy discussion about JackBe and its role as a provider of 'Real-Time Intelligence' software.

Does the definition of Business Intelligence need to evolve, expand, become more nuanced? If these caretakers of BI didn't burn us at the stake, perhaps the entire industry needs to adjust how we think of 'Business Intelligence'.

Here’s a great example: on the 25th anniversary of the data warehouse Colin White caught the attention of many when he blogged “Is a Data Warehouse still required for BI?” When you can pull data directly from authoritative sources - without caching/storing/copying it, instead providing live, ‘real-time’ connectivity – is a data warehouse still necessary? That's some radical thinking.

As a counter point to Colin’s datawarehouseless BI, Barry Devlin (the founding father of data warehousing, no less) brings up a good point in his blog “On Business Intelligence and Real-Time Intelligence”: data warehouses, although they can create delays in intelligence-delivery, regulate the validity of the intelligence produced.

“Does it make logical business sense to combine sources A and B? Does source A contain data from the same timeframe as source C? Does profit margin in source B have the exact same definition as that in source D?”

They both make valid points, so who's right? If I may be so bold as to assess the commentary from genuine experts like Barry and Colin, I think they both are. Is a data warehouse required in MOST instances of a BI implementation? Probably. But in ALL cases? Well…you decide. And that is the most important thing: what was once a ‘must-have’ is now simply a ‘must-consider’. The bedrock of BI isn't what it used to be.

And the data warehouse isn't the only example of the cracks in the traditional BI mindset. The entire spectrum of BI capabilities seems to be under assault by broad social and technology changes. Some experts (or vendors, at least) are questioning the importance of OLAP in the BI landscape. And there’s lots of new ideas and innovations that are being fitted into the BI space: social BI, complex event processing, cloud BI, and yes, even real-time intelligence.

In his recent blog “BI Changes in 2011, But to What?” Andy Mulholland (Global CTO at CapGemini) gives a great summary of the reasons behind this BI metamorphosis. In it he states “Traditional BI frequently leads to the identification of the need to act so the rest tends to follow, where as 'new' BI should be providing a continuous scanning capability showing what and where to focus on and then act.” According to Gartner, a “comprehensive and well-balanced BI platform” includes 12 capabilities:

Andy’s 'new BI' mission statement is a tall order and you have to wonder if the traditional 12-part BI architecture simply isn’t enough to fill it.

With all the evidence laid out on the table, it seems the definition of Business Intelligence is going to evolve one way or another. And so we think it’s time to ask you: What is included in your Dream Team Business Intelligence solution? Does your perfect BI include the old standbys like ETL? A data warehouse? Reporting? Analytics? Must it include new-fangled approaches like Social BI? Self-service? Real-Time?

In short, how do YOU think we should define Business Intelligence? Perhaps the right answer is much simpler than a list of features or a formal architecture. Perhaps BI is simply whatever helps you make the best decisions.