Interactive voice responsive (IVR) systems and their online chatbot counterparts leave a lot to be desired when it comes to customer service. Most people wind up asking to engage with an actual customer service person either because the issue they want addressed can’t be handled by a series of recorded responses or navigating their way through canned responses that have been programmed into a system is simply too frustrating. The good news is that a forthcoming wave of artificial intelligence (AI) technologies promises to dramatically improve what until now has been a less than stellar customer experience.
For example, Google at the recent Google Cloud Next 2018 conference unfurled Contact Center AI, a cloud service available in alpha that provides access to pre-configured AI models that have been optimized for use within IVR and chatbot applications. To make it simpler to design the workflow across an AI-infused contact center application, Google is also making available Dialogflow, a cloud service for building virtual agents using speech algorithms developed by Google. Providers of contact center platforms such as Appian, Chatbase, Cisco, Five9, Genesys, Mitel, Quantiphi, RingCentral, Twilio, UiPath, Upwire and Vonage have all signaled their intent to make use of the Google AI services.
At the Google Cloud Next conference, eBay demonstrated a prototype of how it plans to use virtual agents built using the Genesys contact center platform infused with Google AI to help a customer not only return a pair of shoes, but also refer that customer to a human customer service representative to sell that customer a pair of shoes to replace the ones just returned.
In general, Dan Leiva, vice president of customer service technology for eBay told conference attendees, the biggest challenges with customer services are navigating phone trees, having to provide the same information over again, and the fact that customer service representatives usually need to put a customer on hold while they look up information. AI technologies should go a long way toward mitigating all those issues.https://o1.qnsr.com/log/p.gif?;n=203;c=204663295;s=11915;x=7936;f=201904081034270;u=j;z=TIMESTAMP;a=20410779;e=i
Broader and Deeper Customer Service with AI
Google is clearly not going to be the only provider of these types of AI services. For example, Genesys plans to employ cloud AI services within its platform depending on use cases and which service provides the richest experience at the time an application is being developed, says Chris Connolly, vice president of product marketing for Genesys.
“Our strategy is going to be mix and match,” says Connolly.
AI services from Amazon Web Services (AWS), Apple, Microsoft, IBM and others are all part of the AI customer contact puzzle, says Connolly.
The two primary benefits of those AI services will first come in the form of virtual agents that are capable of handling a much broader range of customer support issues without any human intervention using natural language processing. Beyond that self-service capability, customers will also notice that human customer service representatives have become a lot smarter as well, says Bryan Stokes, senior director of product management for Vonage.
Customer support personnel, when engaging with customers, will be able to in real time to tap into a much richer knowledge base, says Stokes. That’s because AI-infused contact center applications will be able to “listen” to customer interactions. Based on those conversations, the application will then be able to make recommendations concerning how best to resolve a customer issue, says Stokes. The amount of time required to handle those customer issues is also going to decline, notes Stokes.
“Time to resolution is going to drop,” says Stokes.
The customer service agent doesn’t need to become expert in every attribute of a product or service, adds Stokes. That capability should lower the cost of delivering a better customer experience in a support function where staff turnover rates are notoriously high, adds Stokes.
Of course, there’s no magic bullet when it comes to infusing AI into a customer service process. Each company with a vertical industry segment is going to have to train and tune AI models to address unique use cases that only apply to their customer base, says Josh Haslett, vice president of strategic innovation for Mitel. For example, there are specific processes that only apply to customer service involving vacation travel, notes Haslett.
“The AI model needs to understand customer intent,” says Haslett.
Organizations will also need to create the content to train the AI model and then teach that AI model to recognize how to escalate a customer service engagement whenever an exception is encountered, says Haslett.
Virtual agents that make use of AI will also need to be able to remember interactions with customers as they occur across multiple communications channels, adds Haslett.
Only once all those issues are addressed will line of business executives be willing to put their trust in virtual agents to manage anything more complicated than being able to triage inbound customer support calls, says Haslett.
Demonstrating Cost Benefit for New AI Applications
Miriam Hernandez-Kakol, global leader for customer and operations advisory for KPMG, says because organizations have already invested billions of dollars in IVR and other customer service technologies with mixed results, they are naturally going to be cautious when it comes to investing in new applications. Those new applications will need to clearly demonstrate how they either reduce costs or enable an organization to drive additional revenues by increasing customer satisfaction, says Hernandez-Kakol.
KPMG has announced its intent to extend the Intelligent Interactions solution by integrating Contact Center AI services from Google. That offering combines voice, chatbots, virtual assistants, search and email in a single platform developed and serviced by KPMG.
Hernandez-Kakol says one way those AI investments will be justified is by providing a much more personalized experience to the end customers. AI applications, after all, never forget what they learn or need to take a sick day.
“The opportunity for personalization is limitless,” says Hernandez-Kakol.
In fact, there may even come a day when AI-based virtual agents will be able to identify new sales opportunities based on what a customer is inquiring about in real time or has previously purchased, notes Hernandez-Kakol.
Naturally, customer acceptance of AI-infused virtual agents is going to vary widely. For every customer that insists on interacting with a human, there will be another that is perfectly content to never engage with customer support personnel again. The opportunity and challenge now is determining where the fine line between those polar extremes resides for each individual customer.