Using AI To Augment Customer Service Agents

augmented customer supportThere is seldom a sector or field that is not undergoing some form of transformation at the moment, many at the behest of a tidal wave of digital technologies that are upending the traditional way of doing things.  While many of these claims suffer from frothy over-hype, the way companies engage with customers is one domain in which the claims are justified.

A recent report from Deloitte highlighted how customer service has transitioned from the cost center of old, and the most sophisticated companies now aim to create experiences that delight customers and turn them into loyal devotees of the brand.

Of course, for many of us, our experiences engaging with companies are anything but joyful, as technology sends us on a Kafka like dance whose seemingly soul purpose is to prevent us speaking to an actual human being.  This is a prime example of technology being deployed to reduce the costs involved in customer service rather than in actually helping to improve customer service.

A helping hand

Whereas many companies have used technology to replace human(e) customer service, the more enlightened are using to firmly augment the kind of experiences customers receive.  These examples are still using automation, but they aim to empower agents who are often juggling several calls at once.  They help these agents understand the intent of the call or chat, thus helping them to deliver the right answer, more effectively to the customer.

This is possible because humans and AI have fundamentally different skillsets.  For instance, humans can be tremendous at detecting when someone is frustrated, and respond with concern and empathy.  AI-powered systems, by contrast, are great at pulling data from various different systems to help make an almost instant judgement.

One of the more interesting use cases of using technology to help make customer service agents smarter involves machine translation.  For instance, a recent study from Washington University in St. Louis showed the value of machine translation to eBay.

The researchers examined customer service transcripts from the site before and after the introduction of AI-driven machine translation technology in 2014, with a specific aim of exploring the impact on international trade on the site.  They hypothesized that distance is often a barrier to trade because trust is difficult to establish, and therefore being able to successfully communicate will increase trust and improve international trade.

That is indeed what has happened, with even the relatively rudimentary technology available in 2014 seeming to bolster international trade on the site by around 10%.

“These comparisons suggest that the trade-hindering effect of language barriers is of first-order importance,” the authors say. “Improved machine translation has made the eBay world significantly more connected.”

This is driven by the strong desire to solve the customer’s problem as quickly as possible. Machine translation company Unbabel recently published report on the state of customer experience highlighted this as the key challenge company’s are grappling with. The ability to empower agents so that they can both understand the customer’s query and then resolve it as quickly as possible is fundamental to delivering a great customer experience.

“Speed, accuracy and quality of problem resolution are the key expectations of customers today,” Unbabel’s VP of CX Michael Ollitervo Murphy told me at the company’s recent Customer Centric Conference.  “As a deliverer of customer experience, it’s imperative that these things are the fundamental baseline of what you’re offering to consumers.”

Human-machine collaboration

Machine translation is just one of the ways that technology is augmenting the work of human customer service agents.  These technologies are typically making agents considerably smarter by giving them the data they need, when they need it.  For instance, they can provide real-time product information to the agent, so they’re aware of any discounts or if the product in question is out of stock.

What’s more, technology can often perform much of the grunt work that humans are not only not best suited for, but usually don’t enjoy a great deal.  This frees up staff to work on the prickly and creative challenges that humans are perfect for.

The best implementations of AI in customer service are not, therefore, to replace human agents, but to empower them to do a better job at creating the kind of exhilarating customer experiences Deloitte highlight as being key to success in the modern economy.

This marks part of an ongoing trend in which knowledge workers have used digital tools to do their job more effectively, and customer service will be no exception.

“With digital tools we have the ability to make communications really relevant to every customer, and human agents can then ensure that these conversations retain the human touch,” Ollitervo Murphy continues.  “Fast answers that are wrong are no use, but if you can deliver real solutions as quickly as possible then that’s a real game changer.”

With customer experience increasingly regarded as a key competitive battleground, it’s a process that companies can’t really afford to miss out on.  The tools are increasingly available to support the delivery of such exceptional customer experiences, and the challenge now is to ensure employees are sufficiently equipped to utilize them.

Facebooktwitterredditpinterestlinkedinmail