In a recent article over on All Analytics – Analytics-Business Alignment Needs Work – Beth Schultz discussed a set of Gartner predictions written up by Doug Laney. In particular she highlights Gartner’s finding that companies
cited …aligning BI initiatives with corporate strategy and objectives nearly three times as often as they called out technology-related issues
This is a problem I see regularly across both BI and other analytics efforts. I think the primary driver of this is an abject under-definition of the decisions that BI and analytics are supposed to be enabling/improving. We model and manage business processes and the data they need. We don’t model or manage the decisions within those processes nor consider their impact on the business. I call this process Decision Discovery and wrote about it extensively in Decision Management Systems: A Practical Guide to using Business Rules and Predictive Analytics – the whole of chapter 5 is devoted to this critical first step. I wrote a quick overview recently about the use of decision discovery to gather requirements for advanced analytics and my experience is that it also works for driving successful BI efforts. Organizations that understand the decisions involved, how they relate to each other, what information and know-how is needed for each decision and how decision decompose into smaller, more focused ones are more likely to see where analytics will make a difference. By evaluating the impact each decision has on business objectives and key performance indicators, the alignment of analytic efforts to corporate objectives can be achieved – they are linked by the modeled decisions.
In addition, Doug is quoted as saying
Too often we observe organizations developing and deploying hindsight-oriented reports and/or query applications focusing on metrics that users may find interesting, but that don’t represent the operational or strategic controls used to facilitate business performance.
Once again the linkage of decisions to objectives helps as this allows the analytics to be prioritized and focused on assisting with those decisions that really matter to the key objectives of the organization.
In his recommendations Doug has one that particularly resonated with me:
…a top-down view of metrics, which should determine and define your tactics
And this top-down view of metrics should be combined with a business-centric, top down view of decisions.
I also liked
Make experimentation and closed-loop implementations standard practice
As the third step in my approach is always what I call Decision Analysis – the ongoing process of improvement, test and learn and decision performance monitoring.
Decision Management, and the discovery and modeling of decisions in particular, really drives business/analytics alignment.
If you are interested in how to go about decision discovery or what to build you own decision inventory, drop us a line firstname.lastname@example.org and we will see if we can help. You might also want to check out my forthcoming workshop on Business Friendly Data Mining at Predictive Analytics World.