BI versus Decision Support

As an old timer in the industry, I remember the days when the term ‘decision support’ was commonly used in reference to the reporting and analysis capability embedded in IT systems. As all kinds of new fangled ideas then came onto the market, however, carrying much more dynamic labels such as ‘data warehousing’, ‘analytics’ and ‘business intelligence’, the older phrase became unfashionable and ultimately regarded as synonymous with the production of the infamous ‘doorstop report’, containing far more management information than anyone could ever use effectively.

Lately, though, I have been involved in a number of activities around the management and use of information for business decision making, and I have to say that I am starting to miss the old terminology. The trouble is that when people hear the language in common use today, they tend to map the various terms onto technology categories for either the heavy lifting type of extraction, transformation and collation of data often associated with data warehousing, or the heavy BI style number crunching and analysis that typically follows.

When we then see reports from analysts highlighting business intelligence as being high on the CIO agenda, it can all get quite misleading. Our own research at Freeform Dynamics consistently tells us that many organisations continue to struggle with the problems of data fragmentation, duplication, and so on, and indeed this is increasingly an area for improvement that is recognised and prioritised for investment. But in technology terms this encompasses a whole range of information management and delivery solutions beyond classic BI.

One of the things I have been up to recently, for example, is writing a short book on the use of information at the edges of the business, which is basically about surfacing the right data at the right level and the right time in a business process. In many cases, this does not involve a lot of ‘analytics’ per se, as the data required may actually be quite simple – e.g. an aggregate view of customer activity in call centre context. The problem to be solved is therefore around systems integration, rules definition, workflow execution, and so on – basically the coming together of information management with domains such as service oriented architecture (SOA) and business process management (BPM). Get this right, though, especially if you can take an architectural approach to information integration and delivery, and you can enable even relatively unskilled workers on the front line to make key decisions, avoiding the overhead and delay of referrals, escalations, and so on.

Beyond the structured process world, a lot of the challenges among professional workers when comes to getting their hands on the information they require to do their jobs effectively are to do with identifying, locating, extracting and collating both structured and unstructured data from various internal and external sources. That’s not a job for a traditional BI toolset either, that’s about everything from portals, content management and enterprise search, through to collaboration and potential even social media systems.

My suspicion is that when CIO’s are asked about their priorities, and one of the options on the survey form is BI, they take this as proxy for all forms of activity and investments aimed at helping their workforce make better decisions. It’s the old problem of mixing up subjective technology categories with an objective definition of what organisations are trying to achieve. The misleading view thus comes from the round trip – the researcher asks about BI investments, the CIO translates this to spend on helping people make better decisions in general, and the analyst interprets the response as referring to budget allocated to business intelligence solutions – either that, or there are going to be some very surprised and happy BI sales guys out there.

Personally, I would reinstate the term ‘decision support’, as this encapsulates the problem definition very nicely. Sadly, it’s probably got too much baggage to be resurrected, however, so in the meantime, let’s all at least remember its spirit. Enabling effective decision making is not always about the clever number crunching end of the spectrum. In both an operational and professional worker context, huge benefits can often be gained from implementing far less glamorous capability.

While traditional business intelligence continues to evolve into areas of predictive analytics, in-memory databases, and other sexy technology, which are all potentially very important and valuable, it is therefore worth taking a look at the amount of time wasted on the front line of the business, and the risks that arise from a simple lack of basic information access. The chances are that there is some low hanging fruit just there for the taking.