Featured article

The Customer Service Chatbot: What It Can and Can’t Do

We need to talk about something… and that something is the customer service chatbot 🤖.  Chatbot customer service programs have become almost too popular to ignore. But when it comes to customer service, a bot isn’t a one-size-fits-all solution. 

In this blog, we will explore:

  • What a customer service chatbot is
  • Where they fall short 
  • The top customer service chatbot use cases
  • The benefits of using live chat with AI assistance vs. chatbots

What is a Customer Service Chatbot?

A customer service chatbot is a bot that uses machine learning and AI software to answer basic customer questions via live chat. For example, these bots can answer questions like, “how do I reset my password?” or “what are your business hours?” Importantly, these chatbots operate independently. There is no human on the bot side of the conversation. 

Leveraging a chatbot for your customer service program enables you to either immediately answer a customer question or route the question to the most relevant department, resulting in a quick and efficient resolution. But while chatbots can handle multiple conversations at a low cost, they can hurt the customer experience too.

Chatbot Customer Service: Where They Fall Short

According to a report from TalkDesk, “More than half of [survey] respondents (58%) said their customer service expectations are higher today than they were a year ago.” And after live chat became the top communication channel for brands in 2020, 70% of customers say they expect real-time conversation from every business (Zendesk’s 2022 CX Trends Report)

Companies struggling to keep up with these rising expectations often use chatbots to decrease costs while giving customers a service they’ve come to expect. But automated chatbots can leave customers frustrated by often misunderstanding or being unable to provide what the customer is asking for or about, leaving them in an endless queue of follow-up questions that aren’t relevant to their needs.

In fact, 38% of consumers claim poor user experience, and 29% claim frustrating and awkward automated responses are their top live chat complaints. As if that wasn’t enough, 61% of customers say they would switch to a company’s competitor after one bad experience.

But there is a silver lining: 81% of customers say a positive customer service experience makes them more likely to return. That being said, it’s critical to know when a chatbot can handle a situation independently and when a human needs to intervene.

The Best Customer Service Chatbots Know When to Go Away

Organizations using chatbot customer service programs must remember one crucial fact: there’s a limit to what they can handle. While they’re a convenient and low-cost option, chatbots can’t converse like humans — despite current advancements in AI and Natural Language Processing (NPL).  

Human language is often nuanced and emotional. No matter how much time, energy, and resources are dedicated to expanding an NPL model, a chatbot cannot fully read into the emotion a human brings to text conversation (sarcasm, the anger of an all-caps sentence, etc.). The best way to ensure that a customer will have a satisfactory experience is to know when to hand the chat over to a human.

Chatbot Customer Service Use Cases

According to Dan Gingiss, customer service expert and author, “Rather than a replacement for a human agent, though, you should think of chatbots more like a self-service tool or a search engine.” 

Here are some examples of situations that a chatbot can handle itself and situations where a human should enter the conversation:

Chatbot

  • Password reset
  • Request for a refund
  • Appointment scheduling

Human + AI Assist

  • Longer delivery time than promised
  • Hours of operation
  • Poor experience on website/in a store

Now that we know chatbots without a human in the loop should be used as more of a search engine rather than a conversational tool, let’s explore how AI can assist humans in giving customers the best experience possible.

Chatbots vs. Human Agents: How Artificial Intelligence Fits In

Deploying human agents is the only way to give customers the empathy and attention they expect from a service experience. But even with a human in the loop, it’s still possible that emotions and questions can go unnoticed. After all, humans are… only human. 

But bringing together humans and AI assistance to work as a team could be the solution to this problem. 

Slice is a great example of a successful chatbot customer service use case (and an example of how not understanding customer sentiment can be dangerous). 

The online food ordering platform for independent pizzerias used to struggle with customer conversations. Whenever local businesses were backed up on orders during a dinner or lunch rush, Slice agents had to field large spikes in customer conversations. With no support to help chatbots and agents better understand customer sentiment, these situations often led to appeasing angry customers with massive discounts. It simply wasn’t sustainable. 

Slice needed to reduce the duration of customer conversations, ensure that agents adhered to company policies, and reduce appeasement spending.

With the help of live chat with AI assistance (through Loris), Slice solved their problems with ease and was able to:

  • Decrease average handle times by 12%
  • Decrease average CX response times by 20%
  • Increase agent compliance with appeasement policies to 81%
  • Decrease appeasement credit spend by 36% 

Benefits of Live Chat with AI Assist vs. Chatbots

The lesson is this: If you only use a rules-based chatbot, you’re setting up your customers for a sub-par experience. It’s best to use chatbots for simple requests and then deploy humans with AI-assist technology to handle more complex customer conversations.  

Loris does just that. Our tool helps agents make faster, smarter decisions and communicate with customers empathetically by providing intelligent, dynamic language recommendations in real-time. 

Loris also offers comprehensive insights into conversations and customer sentiment, helping agents take real-time action and supervisors provide immediate feedback.

Ready to Make AI Your Ally?

Every customer interaction is an opportunity to create a relationship through real-time responses and a passion for customer satisfaction. 

Book a demo and see how Loris improves agent productivity and customer retention and makes your customer support more human, empathetic, and scalable. 

RECORDED PANEL DISCUSSION

Chatbots Aren’t Enough

Explore when and how to empower your agents.

CSAT Prediction vs Surveys:  Watch our recent panel discussion on the benefit of gaining sentiment scores across 100% of customer interactions.