Intelligent chatbots and AI are the future of webcare
As we said earlier this year, the chatbot era is set to take off in 2018. We have all become familiar with the typical question-answer robots. But the question is: where do AI’s opportunities lie in ‘complex’ customer contact, for example in the event of a complaint, technical support or purchase advice? Consumers often want to be helped well, quickly and personally, but what role does an AI chatbot play in this? And what does the future of webcare look like? Our experts give their view!
The future of webcare: quick and personal contact using chatbots
For some time now we have been seeing a shift from customer contact via public social media channels to customer contact via messaging channels, such as WhatsApp and Facebook Messenger. For example, in an earlier article about webcare at housing corporations, Dafne Verheij said that volume continues to grow every month through various channels, but that 90% of customer contact is handled through WhatsApp. Besides that, live chat is also a very popular channel when it comes to service.
The trend from shifting communication to more closed channels leads to an increase in messages and entails all kinds of follow-up questions, especially as the growing pressure must not be at the expense of the speed and quality of webcare. Let’s get a chatbot to solve the easy questions? Do we opt for channel control, in which we refer customers to the self-service platform? Are we hiring new (webcare) employees? In practice, we see that many organisations are struggling with such issues.
“We see creative solutions to these problems in the marketplace. Bots are already a valuable addition to webcare for many organisations. New technology brings a huge blow to efficiency, but it is precisely the combination with people that ensures success. This makes people’s work more valuable, because they can be used for conversations where their strength and expertise contribute more. Repetitive, simple questions can be captured (partly) by a chatbot. We also see that consumers are satisfied with this: they receive a quick and efficient response”, according to Daniëlle Janssen, Operations Director at OBI4wan.
The perfect mix of people and technology
Whereas for a long time we were ‘afraid’ that the automation of processes through bots would result in less human contact, we can now state that the combination of man and technology will bring success. Showing sympathy and empathy is really only for the human webcare agent. There are enough situations conceivable in which truly human contact can make a difference: from incident management to personal advice.
Nevertheless, a chatbot can be a good addition to the team that now operates on webcare. Besides speed and quality, the application of bots and AI for webcare is particularly interesting because of its efficiency. A clever design of a webcare tool, in combination with a thorough webcare strategy, is often aimed at making the webcare team work as efficiently and effectively as possible. The addition of a chatbot, for example for answering first-line questions, as a work planner for the webcare agent or for taking routine work off your hands, can result in great efficiency gains.
Suggestive conversational agents... What?
Chatbots or Conversational Agents, as we call them, is the way to help the webcare employee in his work. The OBI Bots team ensures that we develop chatbots on a daily basis that learn what the best answer is to a question. This is based on conversations in the past and information systems such as CRM. The answer is presented to the webcare team, after which it can be accepted or modified. The chatbot learns continuously, so the answer given to the question can be used again to modify his trained model. This allows the user to do their work more efficiently and leaves more time for complex questions.
“Our chatbots use AI to understand customer demand and translate it into useful data. The bot learns by analyzing conversations from the past, so that it becomes better and better at recognizing customer demand. In addition, we provide the business rules that he can use to determine what the next steps are. Which systems is the bot allowed to consult? What questions should he ask? And what information does he need to solve customer demand? The chatbot uses AI to cleverly plan the dialogue and not ask unnecessary or clumsy questions. These rules keep your organisation in control,” says Frank Smit, CIO at OBI4wan.
Conversational agents and chatbots, however, are not the only application of Artificial Intelligence. We therefore added the first AI application to OBI4wan: proposed tags. The solution learns from tagged messages from the past what a suitable tag is for a new message. This means that the webcare employee spends less time tagging messages and you maintain an unambiguous structure in the tagging of messages.
The customer chooses the channel and wants to be helped quickly, well and personally. That is the starting point. Every bot or AI application is made to measure. That is why, together with our customers, we are looking for an optimal set-up of a chatbot within the organisation. We see a high added value of data for the development of a successful chatbot. Based on historical data, we scan the messages that appear around your organisation. The scan gives a clear overview of which challenges can be taken away by a chatbot.
“By doing an analysis, you can choose a chatbot based on hard figures, instead of relying on intuition or a fascination with the technology. An analysis of webcare conversations shows what a chatbot delivers in terms of cost savings or reduction in response times for an organisation, but also provides insight into where a chatbot should not be used. Think of a complaint from a customer that requires an empathic response and can therefore be better answered by an employee,” says Koen Hallmann, Data Analyst.
This is the next step in the field of chatbots and webcare automation. With the advent of new techniques, more data and more interaction between systems, bots, FAQs and self-service systems are made smart. The quality of the interaction is paramount, making it possible to better understand the customer’s questions and ultimately to answer them in an efficient way.