Leading customers to the answers they’re looking for is one of the main tasks of customer service. Get that part wrong, and you can quickly find yourself losing customers. But customer service is cost intensive, which is why companies pay attention to efficiency. Nevertheless, their goal is to provide customers a convenient way to get fast and correct responses — in the store, on the phone, or on the web. As an additional support option, chatbots are exceptionally helpful and respond much faster than any human can. But answers from a chatbot are only as good as the data it relies on.
Simple bots work much like search engines. Responding to spoken or typed keywords, they gather information and present it to the user. Bots comb databases that contain questions and the corresponding answers to respond to these simple queries. That’s how they are able, for instance, to quote prices or offer information on specific product features. For this to work, good knowledge management is of course a must.
How bots became chatbots
A while ago, virtual assistants such as IKEA’s Anna and Microsoft’s Clippy started to sneak into our daily lives. By handling simple questions, they were supposed to ease the burden on service agents. But many users became annoyed with the way these digital characters constantly demanded attention. Modern bots converse with customers, which is why they’re called chatbots. With self-learning algorithms and artificial intelligence, chatbots are able to answer complex contextual questions. This capability has opened up new horizons in customer service.
But it’s important that customers understand they’re not talking to a real person. That gives them a chance to adjust their expectations. Which is necessary, because there are limits to what chat bots can do. These digital assistants can however provide very good answers to simple questions like the ones many companies already have listed in their FAQs, such as “How do I return an order?” But any time an answer requires additional information that is stored in a different database, that’s when things get interesting. Answering a question such as “When is my package going to arrive?” requires the chatbot to combine customer information, order status, and shipping information. It is possible to implement that functionality, it just requires the right interfaces and a good algorithm.
What is an intelligent chatbot?
Intelligent chat bots can and should use algorithms that access data stored in connected CRM or ERP systems and other sources to provide answers to spoken or written questions. The more complete and accurate the underlying data, the more precise the chatbot’s answer. But even in the best of cases, you can’t expect miracles. A chatbot can provide competent answers to simple questions from customers, for instance about the company’s products or services. New bots such as Alexa will also gladly accept orders. Companies should use this capability to ease the pressure on service employees. That way, service staff can concentrate on providing empathetic customer care and answering more complex questions. The true strength of chatbots lies at the moment in answering standard questions based on good knowledge management. Wait times for customer service in this case are practically non-existent — with modern technologies, it only takes a chatbot milliseconds to locate data.
Investing in chatbots should be well-planned so that customer satisfaction does not suffer and information security is ensured. The legal and technical challenges involved should not be taken lightly here. Chatbots are also getting smarter all the time with their artificial intelligence. In the future, they will be able to answer more complex questions as well.
Chatbots alongside humans = clear efficiency gains
But often enough, chatbots are still limited and inflexible. They quickly become overwhelmed and cannot help the customer. That’s when a person has to step in. It’s ideal if the machine can seamlessly transfer tasks, because it’s the overall quality of the service that counts in the end. This type of human / machine cooperation can sustainably lower support costs. The call center agent is only called in for complicated cases. The standard questions are answered by the chatbot, which also assimilates new answers provided by agents. This helps the bot learn to solve more complex problems. All historical customer data is immediately accessible to the bot, which allows it to provide personalized responses.
It’s going to take some getting used to before customers completely accept bots. According to the Customer Pulse Survey by Accenture, which surveys some 25,000 customers from 33 countries each year, 73 % of consumers prefer interacting with a human when they require customer service. But respondents expected fast replies from digital channels for standard questions without having to wait. A good customer experience requires collaboration between people and machines.
Say hello to your new coworker
For now, chatbots are simply a new customer service channel. Nothing more, nothing less. They’re not a cure-all, and they’re not about to completely replace humans. But they can considerably simplify the work being done at service centers while enhancing customer satisfaction. For those difficult or even emotional conversations, however, human service representatives are still needed despite the availability of artificial intelligence. Humans working side-by-side with machines is the future. To provide customer service, they both need the same thing: good data. That, in turn, means knowledge management — the only way to turn people and machines into a strong service team — is the future.