Q: BI professionals must stop analysing the past and focus on the future - but what steps do they need to take to achieve this?
A: The most important thing is to begin with the decision in mind - only by understanding the decisions that they are trying to make, and make more accurately, can they make any progress.
Once you know what decisions must be influenced you can begin to ask the question: what would be helpful to know when we make this decision?
Most of the time you need to know what's going to happen next, what the impact of a decision might be and what the context for the decision is. This helps you focus on the future, not the past. Of course you can only predict the future by accurately analysing the past, so it is not as though you can ignore the past. You just have to analyse it in the context of predicting the future.
Q: How important is data quality?
A focus on the future and on predictive analytics also pushes back into data integration, data quality and data management. We need to integrate data, but with a view to making the predictions that will make a difference not just to have integrated data.
[Business intelligence has] the power to improve business decision-making. Everything else is just fluff.
James Taylor, CEO of Decision Management Systems
We need to understand data quality, but that does not mean we necessarily need to fix every problem. Often we can make good enough predictions if we understand the problems, even if we don't fix them. And we need to manage data at a granular level, particularly when it comes to the time dimension. If all we keep are rolled-up totals, we will not be able to extrapolate and forecast the trends that will predict the future.
Q: What kind of decision-making should businesses be targeting?
A: I divide decision-making into strategic - one off, high value decisions; tactical - somewhat repeatable decisions with real value; and operational - high volume, transactional decisions with low individual value and a high cumulative effect.
In general, BI professionals spend too little time thinking about and improving operational decision-making; what offer to make this customer, how to price this deal, whether this transaction is fraudulent... And when they do think about operational decisions, it is often in terms of doing after-the-fact analysis, not the kind of analytics that would improve the decision itself.
Q: Where to start?
All kinds of decisions can and should be improved with analytics and BI, but most businesses should start by focusing on their day-to-day operations where small improvements in decision quality are multiplied by the large number of decisions involved.
Q: What is the real power behind business intelligence?
A: The power to improve business decision-making. Everything else is just fluff.
Q: How is the explosion of unstructured data impacting business decisions?
A: This is an evolving area for sure. Some decisions, many decisions, can be managed using structured data only. But others need to put unstructured data to work. In particular, case management and tactical decision-making often require unstructured analysis.
Today I think the big value comes from extracting structured meaning from unstructured data - identifying the faults that are being discussed in repair notes or the reasons given to collections agents for non-payment. Once extracted, this information can be combined with the original structured data to improve the quality and accuracy of models being developed.
* James Taylor will present a workshop on The Decision-Centric Organisation at ITWeb's BI Summit, on from 23 - 25 February at The Forum, Bryanston. To get more info and book your seat, click here.
Share