There are essentially three areas of focus where AI has made significant strides in automating the customer support domain and is already providing value to enterprises. These areas represent excellent starting points for introducing AI into your organisation and building the intelligent platform of tomorrow. We will be watching these up-and-coming AI trends closely in 2018, and so should you.
Augmentation of human intelligence
The interconnection of humans with technology can be utilised to provide effective decision support during the agent-customer interaction. The agent and the machine collaborate together with the agent’s performance enhanced by the computer’s ability to provide faster resolutions. The bot learns from the agent’s feedback and improves the automated responses over time. This model is especially effective when the contact centre is required to handle large call volumes or highly complex calls. This decision-support tool is expected to reduce agent training time and streamline the entire support process, resulting in a more satisfying customer experience.
Prediction of human behaviour
While AI is not yet able to communicate back and forth with customers, as humans can, there is one area where AI has already surpassed human capabilities: predicting behaviour. AI captures every nuance of human behaviour, whether through chat or voice, and is able to extract customer insights from these structured and unstructured data sets. Comparing data from multiple sources to past patterns can predict sales conversions, customer lifetime value, churn and more.
In cases of technical support, AI can predict the best path to achieve a satisfactory resolution based on a combination of issue type and customer behaviour.
A major driver for the chatbot disappointment in 2017 was the inflated expectations. Based on the hype, companies incorporated chatbot solutions, and expected them to perform on par with human agents within a very short time frame.
While these expectations were not realistic, chatbots as a customer service solution should not be discarded. Chatbots solutions are continuously learning and will continue to improve. They have already been proven effective at handling specific tasks, and over time, will become more proficient at handling the complex tasks as well.
For example, using these bots to automate answers to basic customer questions has been shown to decrease the average agent handle time (AHT) by 10% or more. Companies can also capture the data from the decision-tree based interaction, for later use by a live agent. Clearly, chatbots have the potential to deliver great value.
In summary, to build smart decision-making machines, we must capitalise on the significant progress that has already been made, continue to invest in innovation, take steps to digitise data and identify use cases where even narrow usage of AI in its current form can deliver value to the business. In other words, put the hype and sensationalism aside, and take concrete steps toward long-term success by laying the groundwork for incorporating artificial intelligence in customer service.