What is Cognitive Computing and how does it impact customer experience?
By Guy Mounier, Co-Founder & CEO of CustomerMatrix
What is cognitive computing?
The term ‘Cognitive Computing’ is becoming popular among technology professionals today. It has a very specific meaning when used in conjunction with data analysis. In simple terms, Cognitive Computing is the ability to understand what customers say and do in real-time, and infer what they may be thinking. It involves natural language processing and machine learning to make sense of mostly unstructured data. Mining unstructured data and what is also referred to as ‘big data’ is particularly important, as that is where lessons are learned and best practices reside.
Is it a new concept?
Cognitive Computing can also be compared to a similarity matching algorithm, as it is used to accurately pair current and past interactions in real-time to draw conclusions that will help solve a specific problem or address a need. Incidentally, a cognitive computing engine should also be able to recognize a first-time interaction and not try to give false recommendations in such a context. In theory, Cognitive Computing and Artificial Intelligence (AI) are the same. However, the term ‘Artificial Intelligence’ gained a negative connotation because early attempts to make it work failed due to limited computing resources and data-poor environments. That was 20 years ago. Now, cloud computing technology and Big Data can make AI a possibility, but there is still the need to rebrand it as Cognitive Computing. Often people also confuse data mining with deep machine learning or AI techniques. The fundamental difference is that one technique is static rule-based and one-size-fits-all, while the other is self-learning and contextual.
Impact on customer experience
CRM systems will need a cognitive computing engine to understand customer context in real-time and recommend actions to contact center agents that are most relevant to that specific interaction. If a customer is calling to resolve an on-going technical issue, the system shouldn’t suggest a bundle promotional offer as an action. CRMs need to be smarter about the context. By understanding what customers say and do, and inferring what they think as a result, smart CRMs can help agents build a closer relationship with customers. If call center agents don’t know what their customers want, they will find that recommendations have the same success rate as banner ads. In today’s world, customers don’t want to be interrupted by irrelevant ads. By anticipating their needs or issues, success rates can reach in the high double digits.
It is equally important to know what to do after a customer need or issue has been accurately understood, whether to influence the customer to buy or keep a product/service. By ranking past actions that worked well in that context, this problem can be solved. Often, it is better to recommend that a customer talk to a peer, an expert or send detailed information rather than directly promote a product or service. There has been a shift in the landscape for customer behavior, they don’t want to be sold to but would rather get educated about their choices. In fact, there is a significant risk of destroying customer experience and brand image when pushing the wrong solution to a customer.
Cognitive Computing can help offer predictions and recommendations based on a rational decision making process. But emotions and personality often have a stronger influence than logic on decision making, as demonstrated clearly by the theories of Emergent Behaviors. Based on the writing style of an individual, Cognitive Computing can detect sentiment, opinion about a product or brand and also assess personality types (whether the person takes risks or is risk averse). A good example of a cognitive web service is the IBM Watson Personality Insight. With the help of these extra dimensions, Cognitive Computing offers ultra-precision into understanding not just customer intent but also emotions and personality type, which further helps match them with the right product and service or direct them to the right call center agent.
References
http://insights.som.yale.edu/insights/rethinking-marketing-and-customers-lessons-behavioral-economics
About Guy Mounier
Guy Mounier researched and studied mathematics at Harvard University and MIT, with a specific focus on discovering patterns in "chaotic" non-linear system behaviors and developing algorithms for understanding human-generated content. Mounier's research led him to develop a vision of a new class of smarter software that would amplify the power of people by leveraging IVR, Big Data and cognitive computing to guide users to the best possible decisions and actions, dynamically learn from outcomes, and adapt to real-world situations. Mounier co-founded CustomerMatrix to make this vision a reality and launched the world's first enterprise cognitive system that helps salespeople anticipate customer needs and risks in real-time and recommend what to do next, micro-targeted to each unique situation.