Gone are the days of an annual budgets, plans, and forecasts. Priorities can shift in…
There was a time a few decades ago when the label Business Intelligence made sense. Its original goals were to ensure that the left hand of an enterprise knew what the right hand was doing, all working from the same source of legitimate data.
As corporations rapidly digitized their records and operations, the promise of successful BI was hyped as inevitable rather than just an added expense. Lew Platt, CEO of Hewlett Packard for most of the ‘90s, summed it up aptly when he said, “If only HP knew what HP knows, we would be five times more productive.”
But looking back over the past 20 years, classic BI system failures are legendary. They were difficult to implement and manage, often over-priced and too often became the carefully-guarded fiefdoms of IT or a single C-level officer, shutting out knowledge workers at lower levels. While the market has changed dramatically, with self-service and Cloud-based offerings spurring a tremendous increase in adoptions, there still remain some major design and implementation disconnects that may impede the likelihood that today’s analytics systems will reach their full potential within enterprises. All too often, BI and analytics initiatives are short-sighted in that their goals are defined as providing insights from data feeds, and with that accomplished their implementers think the goal has been reached, when in fact it has not; it’s only the starting point for making a decision.
What business managers need is a flexible tool that continues the journey with them all the way through to the final decision, whose need triggered the search for an intelligence phase to begin with. Then, once that intelligence is provided, you further need a structured decision-making process with governance and traceability of how the decision was made. The decision, often a corrective action, such as offering a customer a deeper discount, or increasing the price of a very popular item, needs to be made by a manager based on the full context and details that led to that point. Traditional BI systems are of little value if they cannot take users on this full journey to decision support, while also capturing the route taken towards a successful decision.
I was gratified to read a Gartner report that agreed with this finding: “The causal link between available data, analytics models and business outcomes is often not understood or articulated. Worse, the recognition that process, application and data need to be reimaged for digital decisions is completely lacking” (How CDOs Engage With Their Stakeholders to Deliver Real Business Value, Gartner, June 15, 2016).
The Gartner report focused on how Chief Data Officers can act as change agents to discover, define and direct business analytics initiatives that will help the enterprise reach tangible business goals through decision-making support. But their findings are equally valuable to IT and business leaders alike with responsibility for upgrading their companies’ analytics systems. By coupling business analytics initiatives with corporate business goals at the very start – thus ensuring that all relevant stakeholders on the business side have a say in what needs to be accomplished with feedback loops to allow course corrections – what had been called BI becomes practical Decision Support processes.
So why don’t we simply retire the term Business Intelligence, or at least admit that it’s become an inaccurate anachronism? What’s far more relevant and accurate a descriptor is Decision Making Support enabled by analytics. The decisions being made may be departmental or C-level, or automated through machine learning and IOT analytics systems, but at the end of the day, technology serves decision making and achieving corporate goals rather than ill-defined BI. Long live Decision-Making Support!