How does one achieve actionable BI?

Business Intelligence or BI by itself is valuable but its full potential is realized only when it is made “actionable” to achieve the desired results in achieving operational efficiencies, customer experience, creating new products/services or transforming the business model.

In my last blogpost, we had looked at some of the fundamental characteristics of actionable Business Intelligence or BI, and what makes it different from the BI reports of yore. In this one, we will look at how organizations can achieve such actionable BI.

The practice of BI, of course, is nothing new. It has been around since at least the 90s. We knew it mostly in the form of detailed reports and dashboards. What many business managers from that era, might also remember is the sense of frustration that one felt when presented with such reports. While they were accurate and detailed, these reports gave us interesting insights about the business trends but making them actionable, required further analysis and processes that generally got lost in organizational silos and politics. By contrast, actionable BI replaces human analysis, decision making and action components with a combination of predictive model(s), applying the intelligence from the models at the point of interaction or for that matter even a pro-active interaction using Natural Language Processing through text (through

chatbots), voice (through voice bots) or other physical action (through robots). This automation of human analysis, decision making and taking actions makes BI fully actionable and increase its value exponentially.

Use of artificial intelligence or AI, which is the critical ingredient in making BI actionable, is becoming prevalent across many domains and industries. The most common example of this kind of pairing of

traditional BI with AI can be seen in the increasing use of Chatbots. It is AI which allows chatbots to interact with customers and interpret what they need. This understanding is then paired with a BI engine that goes through historical data, mines it and comes up with an appropriate and intelligent response. It is thus, the AI component which makes the process actionable, not only in terms of listening, interacting and interpreting a customer’s request but also in terms of formulating the best fit response based on current historical data to predict future behavior

It is estimated that such intelligent chatbots today can answer almost 60-70% of the common queries addressed to call centers. Such chatbots can then perform an initial screening of calls, solve the simpler ones and pass on the appropriate ones to human agents. This can reduce human involvement and bring down costs and response times. Most importantly, such chatbots can lead to optimal decision making and the generation of appropriate responses by applying intelligence at the point of interaction.

But does one need to know AI in order to apply AI? The answer is no, for today, different AI vendors provide a wide variety of Application Programming Interfaces (APIs), which organizations can integrate into their applications depending upon the type of intelligence required by their BI solution. So, for instance, there are a multitude of applications available for image recognition (seeing), translating speech to text (listening), for monitoring activity using cameras (watching), knowledge and intelligence. Any organization can literally put these APIs together with its own knowledge database, to create the required solution.

There are also cases in which clients want a customized AI solution built for their needs and Capgemini/Sogeti can help here as well.

For instance, we recently helped put together a solution for an automobile loan company that helps them initiate pro-active calls and other appropriate actions for customers likely to become delinquent. This solution helped reduce both the delinquency rates and also enhanced customer experience and loyalty. This is an example of actionable BI, which equips organizations with the ability to take pro-active action. This is in contrast to BI in the past, when managers used to get a list of customers who had already become delinquent with no recommendations on the actions needed.

If your organization is interested in exploring an actionable BI solution, you can reach out to me at satya.sachdeva@us.capgemini.com.

I hope this gives you a general framework on how to move towards actionable BI. Please do follow up with me or leave a comment, if you have any specific queries.