Want to learn how to create useful and effective customer journey maps? Join the webinar on Tuesday, April 18 at 1:00 PM ET.
[Estimated read time: 4 minutes]
You: "Okay Google, schedule lunch with Gary for next Tuesday at noon."
Smartphone: "Okay, do you want to save this?"
Smartphone: "Okay, event added to Calendar."
This is a real conversation you could have just had with your smartphone, and is a great example of natural language processing at work. Natural language processing (NLP) refers to technology's ability to understand human language and converse with humans. It can apply to both voice and text-based conversations.
When NLP works, as in the example above, it feels a little like magic. When it doesn't, it can be immensely frustrating -- like when you open a web chat on your favorite retailer's website, and no matter how many different combinations of words you try, you can't get the robot on the other end to give you a straight answer about when your shoes will be shipped!
Customer-facing systems equipped with advanced NLP technology can provide incredible benefits for customer engagement. Seventy percent of customers prefer to self-serve, and NLP powers the ability of self-service tools to interpret questions and provide accurate responses. Customers are more satisfied if they can quickly find the answer they need without calling a 1-800 number and waiting on hold. Plus, fewer calls mean contact center agents have more time to engage customers in meaningful, proactive conversations.
But, when executed poorly, these systems cause the frustration we've all felt when machines don't understand what we're saying to them. That's why it's crucial to the customer experience that NLP works, and that it works extremely well.
In the context of customer interactions with NLP, there are two entities involved in each conversation: the customer and the computer, also known as the conversational agent. The ability of the conversational agent to engage in a two-way dialog with the customer depends on its natural language engine, which accounts for the linguistic features of a conversation. This means the customer and the conversational agent can converse as naturally as possible, as if the customer was talking to another person, not a machine.
On the surface, the process looks fairly linear: text goes in, and text goes out. The customer asks a question or makes a statement, and the conversational agent responds. Or, based on what the customer is doing, the conversational agent can initiate the interaction by saying hello and asking if they need help, just as a store employee would if a customer walked in the door.
However, this process is actually quite complex. How does the conversational agent know what the customer is asking? How does it find an answer to the question? How does it formulate a response the customer can understand?
When the natural language engine receives a customer's question or statement, it breaks it down to its minimal components, such as words, punctuation, numbers, and recognizable linguistic patterns. Next, it builds these components into a specific structure, taking into account misspellings, verb tenses, plurals, and other variations. This structure is used to interpret the customer's root-level intent. The final step is to match a correct response to the customer's intent. A response is generated based on predetermined templates, drawing information from a knowledgebase or other approved source, such your FAQ webpage or a supplier's website. If the customer has a follow-up question or comment, the entire process begins again.
Trends show that today's customers want brand interactions to be quick, convenient, and easy. They don't want to alter their behavior or language to suit what a company or a computer needs from them. This mindset is derived from liquid expectations: If Siri or Cortana or Alexa can understand me and give me what I want, why not you? Never mind that your company (probably) is not Apple or Microsoft or Amazon -- customers expect the same ease and the same intelligence from you as they do of technology giants.
As if rising customer expectations weren't challenging enough, there is also a major disconnect between what companies think they deliver and what customers perceive. For example, 80% of companies say they deliver "superior" customer service, yet only 8% of people think these same companies provide it! Although this statistic may seem disheartening, it presents an opportunity for companies who invest in getting the customer experience right. Brands can drive strong customer loyalty and gain a competitive edge by delivering what customers want: smart self-service interactions that resolve their issues in an easy, natural, conversational way. That's the power of NLP.
Learn how the way consumers prefer to communicate has changed.