Intelligent Enterprise

There was yet another "What's the definition of analytics?" exchange on-line today among some of my industry analyst friends. These debates are typically prompted by a software vendor's claim to be "beyond BI" or the like, as if analytics don't (in my opinion) fall within the scope of business intelligence. Vendor claims of this type are about differentiating on nomenclature rather than on substance, rather than on value delivered to the customer.

My response: "'T ain't what you do, it's the way that you do it." Let's talk value, not feature lists.

Today's exchange was prompted by a vendor briefing hosted by the Boulder Business Intelligence Brain Trust. Search Twitter for the #BBBT hashtag to see the tweet stream generated by attendees. You'll see reports that Vendor X "differentiates analytics from traditional queries & reporting. Analytics are stats, trending, predictive modeling, optimization" and that said vendor believes "93% do not use analytics in their day-to-day jobs." The implication is not only that "traditional queries & reporting" (a.k.a. BI, presumably) aren't enough -- Who'd argue they always are? -- but also that vendors who sell software for "traditional queries & reporting" don't (adequately) provide that other, "beyond BI" good stuff. That is, these word games are about positioning, about ghosting the competition, rather than about substance.

Excuse me for stating what seems obvious, to me at least:

It's time to stop selling (and buying) software on feature lists. This outmoded approach only encourages prospects to create over-reaching "requirements" lists -- long checklists or scorecards of supposedly must-have capabilities -- with too little prioritization or evaluation of the likely business value that can be derived from all those features. Best practices dictate working backward from the desired business outcome to determine the information -- basic numbers or computed indicators -- needed to support decision making, whether automated or with people in the loop. What source data is needed and how can and should it be transformed to produce the intelligence that drives decisions?

These practices and questions capture BI's essential ingredients:

• Information
• Analysis
• Insight-driven business processes

As for "analysis": its totality is not captured in a single set of algorithms or software tools. As implemented at any given organization, analysis may be "rear-view mirror" or predictive, sitting somewhere in a sophistication spectrum that stretches from static reporting to advanced statistical modeling.

BI versus analytics is a false dichotomy. What matters isn't the name, it's that you get the results you need. It's the way that you do it -- the way that you get needed results -- that should dictate the software you choose, the "what you do" of BI-analytics, not the other way around.

E-MAIL | FOLLOW US