Case Study: Customer Engagement with Virtual Cognitive Agents
In its report, "Predicts 2017: Artificial Intelligence," Gartner pegged the number of inquiries about artificial intelligence (AI) through three quarters of 2016 at approximately 2,200. Compare this to three years earlier, when it received virtually no inquiries about AI, and you can clearly see that enterprises are now looking at AI for automating processes that five years ago would have been impossible.
Source: Gartner (November 2016)
While the Gartner inquiries spanned multiple disciplines, the focus of this article is on the use of AI within the contact center environment. Below are highlights of a case study that reveals the opportunities for customer service operators that come with the addition of a virtual cognitive agent (VCA) to Web chat and other channels.
Two years ago, a major energy producer and supplier of oil and gas began researching the use of AI-based tools in 24 global contact centers supporting 3,000 products. This supplier's fuels and lubricants have more than 16,500 characteristics, details of which can be found in a reference library that includes more than 100,000 data sheets.
Customers reach out to these contact centers for many reasons -- for equipment manufacturer recommendations, competitor equivalent products, technical specifications, product characteristics, and complaints, to name a few. In the past, many of these questions required off-line research that, in the worst cases, could have taken days between the time of inquiry and delivery of an answer.
The company selected a VCA platform from Artificial Solutions, a specialist in natural language interaction. Since this company operates in more than 100 countries, language capabilities were high on its list of priorities for AI. Artifical Solutions supports 35 languages, whereas most of the "big" players in conversational AI support fewer than 12.
Using the Artificial Solutions platform, the company began rolling out an AI-based chat bot for its Web chat channel in January 2016. An avatar, tuned by culture and country, puts a "face" to the customer engagement.
Operationally, the energy producer has now consolidated its two dozen contact centers into three locations. My sense is that such consolidation was problematic in the past due to a lack of standardization and language barriers. Conversational AI has a lot to do with solving both problems.
The company started getting measurable results within two months, and by November of 2016 achieved the following:
- 43% reduction in telephone call volume to human agents
- Real-time handling of research requests that previously took hours or days
- 97.4% of customer questions correctly understood by the VCA
- 74% of issues successfully resolved -- and contained -- by the VCA
- 98.8% of VCA answers met or exceeded customer expectation
Consider the last three bullets in light of your current contact center performance. These are profound numbers. Nothing in the customer care business has moved the needle like this since the introduction of IVR in the early '90s.
As I outlined in my previous No Jitter article, text chat is the gateway drug to AI for customer engagement. The opportunity to create efficiency within a single contact center channel is just the start. The ability to expand the VCA's capabilities into other channel interfaces is the next step. Text, voice, and video interfaces can all be enabled.
The use of AI tools to reduce human effort is already having a profound effect on the cost to operate this company's customer support operation. In the long run, these operational savings will be a footnote. The real opportunity here is the reduction in customer effort. We do not have any information related to customer loyalty in this specific case study; however, the opportunity to reduce the effort required by customers in problem resolution will have a positive impact on customer loyalty. In this case, I believe the long-term reduction in marketing expense due to higher customer loyalty will dwarf the operational labor savings by an order of magnitude.