Imagine reading over a movie review. You don’t want to read the whole article, since you want to avoid spoilers, but there are a few key phrases you pick up. “Bad acting” is one of them. “Ludicrous plot holes” is another. Skimming more, you see the unexpected phrase “the worst script of the year.” You didn’t actually read the review, but you know enough to make a decision. You get the gist of the article.

That, at its heart, is language processing. You know the structure of language well enough to decode meaning without having to actually engage with the article. And that goal– understanding context, tone, and nuance – has long been an obstacle for artificial intelligence.

But the field of Natural Language Processing (NLP) has advanced rapidly in a very recent timeframe. This is great news for people managing contact centers and in charge of customer experience (CX) agents. Proper use of this technology will benefit both the customers and the human agents.

NLP won’t replace humans. But it will make their jobs easier, improve morale, and create better organization-wide outcomes.

Speech Recognition: How NLP Works

Right now, you might be asking how Natural Language Processing is different than AI. After all, automated chat boxes have done a pretty good job of reading what a customer writes and helping to prompt them along a preordained journey. But that is more cue-reading than actually processing language.

NLP goes deeper than that. NLP can analyze and understand human language, which allows it to do three things:

1. Interpret speech
2. Measure sentiment
3. Determine which parts are important

What does all this mean? Let’s take the example of the movie, but instead of reading a review, you are listening to a friend who saw the movie and is telling you about it, but is prone to digressions.

“I thought that the movie was, well, I didn’t like the last one the director did, I thought it was pretty trite and boring but in this one he got the right actor, you remember her, she was in the movie about the bees, not not that one, the other bee one, and in this one the script worked really well and the car chases were great.”

That’s the sort of conversation that until very recently would confuse the heck out of robots. But natural language processing helps sort the wheat from the chaff, and helps understand nuance, and would be able to ascertain if your friend actually liked the movie or not. (She did)

Understanding human language is a quantum leap forward. Being able to detect tone is a huge step toward more productive interactions between humans and machines.

Here’s one more example of why. We all have response prompts in our various text or chat apps. If a friend who was nervous about getting a job messaged you “So…I got some news,” your natural response to that ambiguity would be “what is it? How you feeling?” Whereas the chat prompt might be “That’s great!” That’s not a great response.

Better responses come from understanding tone and context. And that understanding is crucial to roles that come into contact with customers.

Understanding the Dual Roles of Natural Language Processing in Customer Contact Roles

Natural language processing in customer contact roles has three distinct roles.

• Understanding customer satisfaction
• Interacting directly with the customer
• Assisting in training CX agents

Let’s break those down.

Understanding Customer Satisfaction

Understanding customer satisfaction has been a (relatively) long-term use of AI and NLP. NLP can analyze, for example, millions of social media posts and gauge not just customer sentiment, but how engaged consumers are with a brand, and what specifically drives their engagement.

For example, one study of the social media engagement posts of 15luxury brands looked at over 3 million posts, and concluded that:

focusing on the entertainment, interaction, and trendiness dimensions of a luxury brand’s social media marketing efforts significantly increases customer engagement, while focusing on the customization dimension does not. The findings have important implications for the design, delivery, and management of social media marketing for luxury brands to engage customers with social media content.

This sort of analysis is predicated on understanding language. Organizations that want to improve their CSAT scores can use NLP to identify areas of dissatisfaction and know where to focus their training.

Interacting Directly with the Customer

Direct consumer interaction is probably the most prevalent use of NLP these days. We are all familiar with chatbots, which can perform such tasks as:

• Onboarding users
• Moving through an FAQ
• Scheduling appointments
• Processing orders
• Tracking orders
• Cross-selling

Of course, the next frontier will be having NLP “agents” actually talking to callers, using their ability to process speech and understand nuance to answer the right questions.

Or course, that is considerably trickier. After all, in a chat there is usually a limited range of questions or responses coming from the human. On the phone, that isn’t as controlled. A person may ask confusing questions, or go off on tangents, or just get frustrated by talking to a machine.

Right now, NLP can help with the simple tasks, the straightforward ones. But people make phone calls when the situation isn’t simple or straightforward. And the truth is, most people prefer to talk to, well, people. They don’t think robots will understand them, and NLP is not yet advanced enough to change their mind.

So that’s where people come in.

How NLP assists in training CX agents

In some minds, there is an inherent conflict between Natural Language Processing and human contact center agents. There has been talk for years about how AI would make call centers obsolete. But the stubborn truth remains that people are simply better at dealing with people.

NLP shouldn’t replace humans –- it should assist them.

We covered one way NLP assists human CX agents, by handling the simple stuff so that the professionals aren’t spending time answering the same questions over and over. Not only is that wasteful, but it is boring and can depress morale.

The best way to use NLP is in training. The thinking is that NLP can detect voice and tone in customers – but it can also do so in agents.

Is the agent being empathetic?

Do they seem confident?

Are they directing the call clearly and concisely?

Do they seem like problem solvers, or are they just reading from the script?

NLP can be used to answer these questions, which can give managers clues as to where to support their employees. If they are having trouble adjusting to changing circumstances, then they need more training on variable scenarios.

Natural language processing is part of the AI coaching platform created by Zenarate. The Zenarate immersive learning platform replicates customer scenarios in a risk-free environment, with the platform being able to analyze, interpret, and give feedback in real-time. This way, the agents and their managers can understand where to improve.

We don’t know how sophisticated NLP is going to become. But we know it can be used to make everyday contact center experiences that much better.

Contact us today to schedule a demo to learn more about how you can incorporate Zenarate AI Simulation Training into your agent training program. We will answer your questions and show you how you can help your organization develop confidently prepared agents while delivering exceptional experiences to the ones that matter most – your customers.