Cognitive analytics is offering organisations the chance to out think competitors. Leveraging new ways of interaction between humans and machines, cognitive systems can use natural language understanding and expert training to unveil new connections and insights within both structured and unstructured data – and get smarter during the process. But where does this new wave of analytics fit in today’s business world?
Organisations are desperate to democratise data and empower business people to better innovate, collaborate and communicate. Cognitive offers huge potential for organisations to transform the speed of decision making. But it's important to remember that cognitive is a complementary, not replacement component of the analytics toolset and, as such, has a very specific role to play.
What cognitive is. There are three words that sum up cognitive analytics: fast, focused and dispassionate.
Cognitive solutions leverage intelligent computing power and analytics to deliver new insights from specific data sets – and they do so without human interaction. In contrast to traditional analytics projects, which involve requirements definition, assessment of data needs, the creation of a data structure and queries designed to meet those requirements, the cognitive process is very simple. Cognitive analytics projects essentially take an existing data set and use analytics techniques to deliver new views of that data. It is quick; focused on a specific business area; and, without any of the human tendency to pre-judge outcomes, dispassionate.
Cognitive is about empowering business users with new connections, from customer behaviour to merchandising, driver performance to patient diagnosis. The value of a dispassionate approach was demonstrated clearly by an exercise undertaken using IBM’s Watson cognitive engine. Analysis of patient records revealed a new set of opportunities for diagnosis that doctors – trained to look at very specific areas – had not considered. The ability to empower business users to apply this thinking to specific operational areas offers huge potential value.
Critically, the scale of cognitive analytics projects is far smaller than the big data and traditional business intelligence projects that have demanded significant investment in both time and resources. As such, while the insights provided by this new view of data offer the potential to transform business operations, the process itself is far from disruptive.
What cognitive is not. Cognitive analytics is not a replacement for existing operational data and Key Performance Indicators (KPIs).
Organisations still need that essential – and hard won – foundation of consolidated, accurate information upon which to base operations. However compelling the results from cognitive systems, no business will operate effectively if the entire company is focused on a mass of new insights without some controlling process in place.
Business performance and decision making has been significantly enhanced in recent years by the provision of a single, trusted set of numbers shared by every team from finance to sales, logistics to customer service. No business should even consider turning its back on this essential information resource. Cognitive analytics must run alongside these existing, trusted data analytics platforms and be used in a different way. Cognitive is about providing business users with a rapid, iterative way of gaining new insights from existing data sets. And while offering massive potential to reveal new opportunities, no organisation is going to risk plugging cognitive analytics directly into systems, from merchandise to customer service, just yet.
Cognitive analytics is compelling – when approached in the right way. By removing human pre-judgement and side-stepping the time and complexity of traditional analytics projects, cognitive is a viable option for companies of any size. It is time to take a fresh look at business data.
By Peter Ruffley, chairman, Zizo
