Amazon Moves Alexa into the Contact Center
Amazon has a reputation for not only delivering excellent customer service but also continually innovating to make the experience better and easier for customers. Now, following Amazon Web Service’s late March release of Amazon Connect, a suite of cloud-based contact center services, companies can quite literally borrow a page from Amazon’s playbook.
Amazon Connect leverages the artificial intelligence powering the online retail giant’s Alexa virtual assistant to handle contact center calls and texts. It also includes new tools, such as Lex, an artificial intelligence–based service that relies on the same automatic speech recognition (ASR) technology and natural language understanding (NLU) that powers Alexa. With Amazon Connect, customers can set up and configure a virtual contact center in minutes. A self-service graphical interface helps users design contact flows, manage agents, and track performance metrics.
It’s unusual for a technology to take root in consumers’ homes and then transition into the enterprise, but for Amazon, this is a logical progression with Amazon Connect and Alexa. Alexa is the engine that drives the personal assistant applications on Amazon Echo, Echo Dot, Tap, and Fire devices. The virtual assistant can be used by consumers for tasks ranging from getting sports scores to ordering a pizza, from turning up the lights in the living room to scheduling a ride from Uber, through thousands of “skills” that are being built for Alexa.
“Amazon has always been about putting the consumer first to deliver great customer service,” says customer service expert and author Shep Hyken. “It makes sense that Alexa started out as a tool for customers, and it’ll remain a tool for customers, even as companies put it to use in call centers.”
“Ten years ago, we made the decision to build our own customer contact center technology from scratch because legacy solutions did not provide the scale, cost structure, and features we needed to deliver excellent customer service for our customers around the world,” said Tom Weiland, vice president of worldwide customer service at Amazon, in a statement. “This choice has been a differentiator for us, as it is used today by our agents around the world in the millions of interactions they have with our customers. We’re excited to offer this technology to customers as an AWS service—with all of the simplicity, flexibility, reliability, and cost-effectiveness of the cloud.”
Experts predict that Amazon’s technology will likely take over some of the basic functions that call center agents now perform. For example, rather than waiting on hold to speak to an agent, customers will be able to get information about their orders, receive answers to billing questions, and solve other basic problems through Amazon Connect. The technology will go further than traditional interactive voice response technology, however.
At its core, Alexa is a document-retrieval tool, but its artificial intelligence layer enables it to learn—that’s what makes it fundamentally different from IVR and other automated response systems. “Alexa’s ability to learn makes the technology powerful,” says Cindy Zhou, vice president and principal analyst at Constellation Research. “She can understand questions and give intelligent responses. She can also make decisions about who to transfer a call to. The experience is much more than just pressing one for billing or two for tech support. It can be conversational and more of a give-and-take than what a customer might get from a typical IVR interaction.”
The Amazon Connect contact center suite includes dynamic skills-based call routing and real-time historical analysis of calls, providing a history of customers’ interactions with the company to determine why they’re calling and how best to handle their calls.
Alexa’s abilities are powerful, but Hyken says the technology is in its infancy, and there’s still plenty of work to be done. Alexa is not as conversational or as contextually aware as Google Assistant, for example.
Once the technology is rolled out in contact centers, it might be able to do more than just handle service requests. It might eventually listen in on conversations between agents and customers to provide recommendations on how to best handle issues, pull up relevant documents and articles instantly, and improve agent-customer engagements in other ways. “We’re talking about a technology that, similar to [IBM’s] Watson on Jeopardy!, can pull up information instantly, faster than a human. That has incredible potential when you think about [the current state of knowledge management technology],” Hyken says.
As for concerns about Connect replacing customer service agents, Zhou says it’s unlikely. Though it might automate some functions, people will still be needed for more nuanced and complicated customer service requests. Hyken agrees, pointing to an old song by the Buggles. “Did video kill the radio star? Not really. Things change, as they likely will in the contact center, but Alexa will not replace all of the functions that humans can perform,” he says.
Amazon Connect integrates with a broad set of other Amazon Web Services tools and infrastructure so customers can record calls in Amazon Simple Storage Service (Amazon S3); use Amazon Kinesis to stream contact center metrics data to Amazon S3, Amazon Redshift, or an external data warehouse; use Amazon QuickSight for data visualization and analytics; and use AWS Directory Service to allow agents to log in to Amazon Connect with their corporate credentials. Amazon Connect also integrates with leading CRM, workforce management, analytics, and help desk offerings from Appian, Calabrio, CRMNEXT, Freshdesk, Paxata, Pentaho, Pindrop, Salesforce.com, SugarCRM, Tableau, Twilio, VoiceBase, Zendesk, and Zoho.