James Kobielus of Forrester brought the concepts of predictive analytics to processes to discuss optimizing processes using the Next Best Action (NBA): using analytics and predictive models to figure out what you should do next in a process in order to optimize customer-facing processes.
As we heard in this morning’s keynote, agility is mandatory not just for competitive differentiation, for but basic business survival. This is especially true for customer-facing processes: since customer relationships are fragile and customer satisfaction is dynamic, the processes need to be highly agile. Customer happiness metrics need to be built into process design, since customer (un)happiness can be broadcast via social media in a heartbeat. According to Kobielus, if you have the right data and can analyze it appropriately, you can figure out what a customer needs to experience in order to maximize their satisfaction and maximizing your profits.
Business agility is all about converging process, data, rules and analytics. Instead of static business processes, historical business intelligence and business rules silos, we need to have real-time business Intelligence, dynamic processes, and advanced analytics and rules that guide and automate processes. It’s all about business processes, but processes infused with agile intelligence. This has become a huge field of study (and implementation) in customer-facing scenarios, where data mining and behavioral studies are used to create predictive models on what the next best action is for a specific customer, given their past behavior as your customer, and even social media sentiment analysis.
He walked through a number of NBA case studies, including auto-generating offers based on a customer’s portal behavior in retail; tying together multichannel customer communications in telecom; and personalizing cross-channel customer interactions in financial services. These are based on coupling front and back-office processes with predictive analytics and rules, while automating the creation of the predictive models so that they are constantly fine-tuned without human intervention.