by Zeus Kerravala

Practical AI in the contact center starts with agent assistance

Mar 05, 20196 mins
Artificial IntelligenceCloud Computing

Many businesses are turning to artificial intelligence (AI) to improve customer experience. CIOs should be careful with the emerging technology, however, and use it as an assistive tool first to reduce risk.

7 artificial intelligence
Credit: Getty Images

A few years ago, Walker and a couple of other firms predicted that customer experience (CX) would become the top brand differentiator by 2020. This gave CIOs a bit of time to plan how to ensure a top-quality customer experience. Based on other research I have seen and conversations with C-level executives, 2020 came early and CX is already the top differentiator. 

Great experiences drive loyalty up, bad experiences drive customers away

CIOs need to understand that CX improvements are driven by new technology, particularly in the contact center, as this is often the first point of communication between a customer and a business when there is an inquiry or a problem. Have a problem with your flight? Start chatting with your airline’s contact center. There’s an issue with your credit card? Call your bank’s contact center. The product you ordered online is broken? Send a text message to the retailer’s contact center.

Each and every interaction with a customer through the contact center will either drive loyalty up or chase the customer away. This is one reason why so many businesses have been moving their contact centers to the cloud over the past few years. It’s the fastest and easiest way to modernize it.

[ Learn the 10 signs you’re ready for AI — but might not succeed. | Get the latest insights with our CIO Daily newsletter. ]

One of the latest trends in the contact center has been the infusion of artificial intelligence (AI). Contact centers contain massive amounts of information, and AI can connect the dots between the data points and change the way businesses interact with their customers. In particular, there has been tremendous interest in using AI-based virtual agents that, in theory, would be indistinguishable from a real person. For example, last year at GoogleNext 2018, there were many demonstrations of people ordering pizza and completing other activities through a virtual agent.

AI is not ready to be customer facing

The thought of having AI-powered virtual agents talk to customers and quickly handle customer calls might sound appealing, but it’s the wrong first approach for AI in the contact center. That’s because virtual agents are still unproven. Anyone who has dealt with a poorly designed chatbot can attest to the fact that they often spew out recommendations or comments that don’t make sense. Customers do not like this and could be driven away by it.

Should an organization decide to leverage a chatbot, it should be thoughtfully integrated into the customer’s experience, handle the highly transactional or easily predicted tasks, and enable an easy and seamless handoff to a live agent if necessary.

Using AI to make human agents better

I recently came across a post by Talkdesk’s Patrick Russell in which he discusses why human agents are still better than bots or virtual agents. It got me thinking that the most practical use for AI in the contact center today is to improve agent efficiency.

A good analogy is to consider what’s happening with AI in cars. The short-term AI plans for most car companies isn’t to roll out fleets of fully autonomous vehicles. Instead, they are using AI to make drivers better by alerting them if they veer outside the lane or take their eyes off the road too long. AI is a tool to improve the driver, not to take over.

Similarly, in the contact center, practical AI revolves around making the agent smarter and more effective by infusing AI into every conversation — but keeping it hidden from the customer. For example, if a customer calls in and has a complaint about a product, AI could share three or four possible responses. If none makes sense, there’s no harm to the customer interaction because the agent will ignore the responses and suggest something more appropriate.

Another example is when a customer calls, AI could aggregate all current and historical information about the person and present the agent with the most likely reason for contact. The fact is most agents initially have no clue why a customer is calling, but an AI-based system could connect the dots and make an agent smarter.

AI helps behind the curtain

There are many behind-the-scenes use cases for AI, as well. Contact center managers spend hours developing complex call routing algorithms to route conversations to the “best” agent. The challenge with this is keeping things up to date. Many modernized contact centers have AI-based call routing capabilities where the machines will put those routing plans together for the company. If the business isn’t ready to make the jump, AI can be used to at least test the call routing plan and then point out where there are holes or make recommendations of things to change.

CX has never been more important to the survival of businesses. CIOs need to be aggressive with AI as an enabler of new experiences, but they must use it in a way that doesn’t put the company at risk. AI-based virtual agents will eventually be ready to roll out to customers, but there’s too much risk today. A practical approach is to use AI to make human agents smarter and more efficient.

More on AI and machine learning: