Shopper Experience: cómo un marco para IA puede ayudar a mejorar la CX
Author: Ankur Garg
According to Olive Huang, research vice president at Gartner, “By 2021, 15% of all customer service interactions will be completely handled by AI, an increase of 400% from 2017.” Furthermore, Gartner declared «customer analytics and continuous experience» as its No. 1 focus area in the CX space.
Over the last few years, the industry has experienced a surge in the adoption of virtual customer assistants (VCA), or chatbots, for customer support. At last year’s Gartner Customer Experience summit in Tokyo, the research firm predicted that over 25% of customer service and support operations would integrate a VCA by 2020.
This adoption of chatbots was followed by «case (or ticket) deflection,» which provided a self-service interface to customers for answering their own queries. Some of the more evolved enterprises started using search engines to provide real-time results to customer queries. A handful of enterprises have now started experimenting with artificial intelligence (AI)-driven searches to pursue the next milestone in case deflection.
The question now is, what’s next? Are there more game-changing methods on the horizon that will use AI to further improve the customer experience? Are we really using AI to its full capacity via chatbots and case deflection, or there is more to it?
After over a year of research, with the support of my team), I was able to create a starter framework for how AI can be applied in a holistic way to enhance customer experience. I call it AURA.
What Is AURA?
Based on the definition provided by Google aura «is the distinctive atmosphere or quality that seems to surround and be generated by a person, thing, or place.»
In the context of customer experience, let’s assume that the «thing» is an enterprise. Obviously, the word aura isn’t typically associated with a CX framework, so I tried to articulate key elements of this new framework with a careful choice of words to arrive at this name.
What Is AURA For Enterprise CX?
It is an acronym I created that means «anticipate, understand, resolve and assist.» I’ll now go over each word and explain how it contributes to AURA and how you can use this framework with AI to improve the customer experience.
To me, this element of CX did not occur until I was deep into case deflection. We already had various methods that we were using, but I still felt that there was more to case deflection than what we had achieved. Well, one day while in a boardroom, when we were mapping out the whole journey of a customer, it suddenly occurred to me: Why are we leaving so much customer data outside of our deflection engine?
Real resolution of customer issues starts when they face an issue, so it’s critical to know as many issues as possible that your customers may face even before they become a customer. This is why anticipation is a key element in this framework. It’s ultimately where you leverage all customer data — purchase, browsing, acquisition, usage logs, persona, demography — that you have got on your customers (or future customers) and forecast their behavior.
Anticipation helps you stay ahead of your customer needs and be ready for them when they need you. These forecasts can be used to plan workloads and human resources.
The second key element of AURA is understanding user needs when they provide you with an explicit act of intent (i.e., chat messages, emails, voice calls, web forms). This falls in the category of intent recognition; however, you can make it much better compared to conventional methods if you factor in the larger user context you obtained from employing anticipation.
Knowing what a user is asking for is still a less solved problem in classical chatbots space. Though it may not seem like it, it is very difficult to configure «intent recognition» systems for nuances of human expression (we as humans do get creative with our expressions all the time).
Hence, augmentation of anticipated context is essential to helping you understand users’ intent toward a resolution to their queries.
After understanding or anticipating, we essentially want to resolve users issues. The preferred (aspirational) method is self- or automated resolution. This saves time, scales much better and increases CX.
This element of the AURA framework enables the self-resolution of customer issues and can be seen as an enhanced version of case deflection. The AURA framework, unlike present-day search-based (or AI search) case deflection, takes self-resolution to the next level by combining user context from anticipation and a better understanding of user issues.
Finally, AI is still nowhere close to being able to perform up to the level of humans, so there will be cases in which need human assistance is required. However, other key elements of AURA can be leveraged to aid humans in performing tasks better and faster as well. Something as simple as an automated case dossier can help a human assistant.
Hopefully, this 30,000-foot view of AURA gave you a basic understanding of the framework and how AI can be used to enhance the customer experience. In future articles, I plan to delve into how each element can be put to greater use.