Chatbots are here and they’re already talking to your customers. Is it the conversation you want them to have? It’s a question you’ll need to answer in the not-too-distant future, if you haven’t already. Yet, chatbots have been here for a while, the difference today is the ease and scale in which they can be deployed and the advances in technology supporting them, such as machine learning and artificial intelligence.

In Australia, one of the first major companies to embrace the technology was banking giant, NAB, which released a chatbot in late 2017. This chatbot, billed as a ‘digital virtual banker,’ is designed to respond to 200 common business banking questions, with a further 13,000 variations of those questions able to be handled by the AI behind the bot.

“If AI can beat humans at chess and Go, then its ability to answer unstructured questions, and revolutionise customer service industries, is just a matter of time.”

If the NAB chatbot can’t handle the customer query, it’s routed through to a human banker.

Customer satisfaction
One of the real questions facing marketers is whether customers are happy talking to a digital agent. Marketers have made technology investments to retain customers, understanding that it’s easier to up-sell an existing customer than it is to acquire a new one. If customers, particularly in high churn industries such as financial services and retail, are not happy with the service they get from a bot or electronic agent, then it’s likely they’ll take their business elsewhere.

Fortunately, the early research indicates that customers are happy using bots – so long as the interaction is a successful one, and they save time.

According to research from bot maker SnatchBot, only 14 percent of chatbot conversations end up with a human and only 12 percent of conversations with an electronic agent are abandoned. In other words, they appear to work.

Better news for marketers is that 80 percent of conversations between a human and a bot ended with a ‘thank you’, indicating the customer was happy with the interaction and was able to get the required information.

For NAB, the number one reason it introduced chatbots was because its business customers are time-poor. Using an electronic agent allows these customers to spend more time on their business, and less running through mundane, long-winded conversations with a human agent to get basic tasks done.

How to chatbot
While it’s still early days for bot technology, the functionality has come a long way. Today, successful chatbots should deliver a series of outcomes including a rich, two-way interaction that offers as much information as possible, backed by human support for the best possible customer and economic outcomes.

Bots should provide better or more accurate advice to customers, while feeding back insights into the business that can be used to enhance its relationship with customers. And those interactions shouldn’t stop there: they should integrate into an omni-channel management solution for further analysis, audit and quality control purposes.

Companies looking to deploy chatbots need to keep several things in mind for a successful strategy. As bots are customer facing, they must be deeply integrated into existing CRM and CCM systems. Fully understand the customer journey, and before rolling out a bot, assemble a multi-stakeholder team to evaluate how a robot advisor can enhance the customer experience.

This team can then consider how the organisation’s knowledge-based offerings can be enhanced, and costs reduced by using bot-provided cognitive services which are enhanced by the human factor.

Broad bot strokes
Now, chatbots work well where the questions asked fall into broad categories, such as “what is my account balance,” and “when is my policy renewal due.” This is why the financial services industry is rushing to embrace the technology, because as NAB has demonstrated, around 200 questions, with associated variations, are enough to answer the most basic and routine queries.

Where chatbots still have a long way to go is addressing unstructured questions or those which require a greater level of contextual awareness. With technology rapidly advancing, and AI being the focus of vast research efforts from the likes of Google, Apple, Amazon and Microsoft, it’s likely that chatbots will become far more sophisticated over time.

If AI can beat humans at chess and Go, then its ability to answer unstructured questions, and revolutionise customer service industries, is just a matter of time.

ABOUT AARON EVERINGHAM//

Aaron Everingham is sales director, Australia and New Zealand, at Quadient