How chatbots can help answer customer questions

Companies are increasingly using chatbots to answer customer questions and improve CX. Pairing AI with human agents gives customers better, faster answers.

Author
Date published
November 02, 2018 Categories

Old school deflection strategies are every customer’s worst nightmare. When on the receiving end of automated email replies or never-ending phone menus, the experience is frustrating, if not infuriating.

Adding insult to injury, many businesses believe that deflecting customer questions away from live agents and into digital channels is risky business, fraught with negative perceptions or worse, public backlash. But the truth is, today’s customers just want answers and they want them fast — 53% of customers prefer online chat over calling, and 67% of customers hang up the phone in frustration when they cannot reach a customer service representative and are left with no alternative path to resolution.

But deflection doesn’t have to be a dirty word. At the end of the day, it’s all about the execution. While getting answers wasn’t easy in legacy phone-based systems, “deflection” to self-serve channels like chatbots and carefully crafted knowledge bases offer a path to quicker, more satisfying resolutions for customers.

From ELIZA to Alexa

One of the best modern-day tools connecting customers with the information they seek are chatbots. Chatbots have come a long way from ELIZA circa 1966 that mimicked human conversation by matching user prompts to scripted responses. Once a rudimentary solution that struggled to accurately respond to customer inquiries, today’s bots are smart, continuously learning and in some cases, hard to distinguish from a human agent.

But first, let’s move past the assumption that customers prefer dealing with real human beings. While it’s true that many still do, an increasing percentage of the population are digital natives demonstrating a willingness to embrace virtual assistants and chatbots. As of 2013, nearly 96% of American millennials were digital natives. Look at the overwhelming adoption of Siri, Alexa and Google Assistant and it’s clear that virtual assistants and chatbots are becoming ubiquitous. Today’s consumers care more about getting answers than they do about how to go about it. Gartner predicts that by 2020, customers will manage 85 percent of their relationship with an enterprise without interacting with a human.

In fact, researchers at CEB have authored the book and coined the phrase “The Effortless Experience” offering evidence that customers don’t want “dazzle and delight,” but instead “want their problems solved as quickly and effortlessly as possible so that they can get on with their lives.”

With lots of conflicting accounts of the good, bad and downright ugly role of chatbots and artificial intelligence, coupled with a growing number of chatbot vendors, it can be tough for any business to see a clear path forward.

Four key takeaways about chatbots and what to expect when you wade in

1. AI-powered bots can live up to customer service standards. Using Natural Language Processing (NLP) and machine learning capabilities, today’s chatbots are able to manage many of the mundane tasks that tie up human agents—whether that’s checking an account balance, paying a bill, resetting a password or helping choose the best shipping alternative. Because chatbots can understand and execute on simple, repetitive tasks, live agents can “deflect” a customer request to a chatbot that can quickly provide the answer the customer needs.

2. Pairing live agents with AI frees agents to focus on tougher conversations. Use a combination of letting your agents take over bot chats or giving visitors the option to talk to a human whenever they want, so agents can focus on the tougher conversations while customers get their routine questions answered at warp speed. Channeling the legendary sci-fi author Isaac Asimov, who devised the rule that no robot shall ever harm a human, in the event a bot chat starts to go sideways, AI-powered sentiment analysis can alert agents to jump into the conversation before the experience sours.

3. AI-enabled chatbots can capture data from customer interactions to better understand and anticipate customer questions. What types of questions are customers asking and at what frequency? From that, companies continue improving customer experience and better determine which requests can be handled by bots versus live agents. For example, if customers frequently reach out to ask about hours, pricing or product availability, companies can ensure their chatbots are equipped to field those questions to free up live agents to focus on the more complex inquiries.

4. Continue to monitor live chat performance and customer satisfaction. Use this data to identify how chatbots and live agents can work better together. At the end of the day, live agents and bots should always be held to the same customer service standards.

A few helpful metrics to keep customer service standards high

1. Total number of chats: By comparing how many chats chatbots are picking up and completing compared to agents, companies can gain insight into how their chatbot is reducing agent workload and saving the company time and money

2. Wait time: Chatbots can help keep user wait time low for conversations with live agents and can use data gleaned from previous chats to quickly determine if a request needs to be routed to a live agent

3. Customer satisfaction rate: Look at how users rate their interactions with chatbots against those with live agents to see how chatbots stack up against the service live agents are providing. They should both be achieving high customer satisfaction rates

Deflecting some customer questions to chatbots

It’s important for companies to know that they can’t flip a switch and have a productive chatbot on staff. It takes time and commitment and a comprehensive training program to get a good bot going (although not nearly as much as it would with a human agent). Once built, it’s wise to ease into the bot world, limiting engagement to a small segment of your customer base. A graduated entry into the real-world breeds confidence that the bot is delivering the expected service level before going prime time.

Once up and running, “deflection” to bots will be the win-win that organizations have been waiting for; better customer experience, a more efficient and satisfied workforce and lower overall cost of service.

Jeff Epstein is VP of Product at Comm100.

Exit mobile version