Decision management expert and consultant James Taylor was 'cornered' recently at the IBM IOD conference and asked to explain the present and future of business analytics. An eloquent speaker and veteran in this field, James does a great job of highlighting the current growth and energy in this space, some of the confusion this has engendered, and the questions you should be asking yourself in determining whether analytics are right for you:
He also highlights three critical issues on which I'll share my viewpoint:
Where can analytics be employed?
Decisions take place across the organization: from the CEO deciding who to appoint as the new sales director, down to the customer service rep asking if you want to take out a maintenance plan when you buy a new computer. At not all points does it make sense to employ analytics to inform the decision process. If you have highly automated business processes in your organization, then company-wide business analytics may make sense. Alternatively, it may just make sense to use analytics to sharpen up one department such as the marketing operation.
There is a political dimension to this which also has to be considered. It could be that the marketing department has a tight agency relationship who strongly pitch an analytics solution highly tailored for the marketing operation. Whilst this may be able to drive up efficiencies in marketing, it won't help the support decision process (or possibly cross-sell or up-sell opportunities).
On the other hand, it could be that the IT department, in the interests of cost-cutting, prefer to go with a centralized solution with a narrower maintenance footprint.
These considerations (which tend to be aggravated in larger organizations) need to be taken into account in addition to the theoretical/modeling questions:
"What are the decisions that drive my business? how do I apply analytics to make a better decision or drive my metrics in the direction I want?"
Figure out how to align Business, IT and Analytics
In the past it was tough enough to engage business and IT departments (which can be heavily siloed and have the kind of relationship you see between Siamese Fighting Fish). But now you’re throwing an analytics team into the tank.
Although I'd suggest that this analytics team can provide the glue that holds together those creative types in marketing and the IT logicians. It's not unusual to find the analytics practitioners sitting somewhere in a department such as the corporate office. Whilst they may have strong knowledge of the tools, they are also plugged into the business imperative. As long as the importance of their role is realized and they are given due authority, they may well be able to spearhead the implementation of a business analytics solution and its systematic application.
Begin with the decision in mind
Why? So you don't end up drowning in data that will do nothing to drive your business. James points out that you need to understand what is a good and bad decision (for instance be clear on what a positive or negative outcome looks like).
Just like the scientist needs to understand there is inherent bias in the questions she asks, so you should realize that the decisions you choose to focus on can have a profound effect on your business. For instance just applying analytics to short term decision making (such as maximizing quarterly sales) could pull you out of synch with any strategic objectives and hurt you in future years. If you go overboard using predictive analytics to decide what to offer individual customers next on your website based on their past behavior, you may end up looking like a creepy stalker. Keep an eye out for symptoms of unintended consequences!
James is one of the most prominent/prolific bloggers in the decision management space and can be found at JT on EDM and ebizQ.
In the video James references these IBM technologies: Cognos business intelligence, predictive modeling from SPSS, and business rules and optimization solutions from ILOG.