Call centre AI needs knowledge

It's near impossible to browse an IT journal without landing on the latest story of how chatbots, artificial intelligence (AI) and machine learning are going to reshape our future. This trend is mirrored in the focus placed on automation in the contact centre space by industry analysts such as Gartner, which predict this technology will become mainstream in the next two to five years.

While many new technologies embed themselves into industry standards gradually, being rolled out with the next version update after cresting Gartner's hype curve, automation may well be the one to buck this trend and be a much-needed disruptor in the contact centre industry, according to Jed Hewson, co-founder and joint CEO of 1Stream.

When we think about what automation can do for a contact centre, our minds go immediately to chatbots, an automated chat system that can provide quick and efficient solutions to a customer's query. But, more than this, automation can take the form of a computerised system that can 'listen' to contact centre interactions and assess them, providing a customer satisfaction score or flagging them for further evaluation. An agent's role can also be enhanced through automation by being able to enter the reason for a customer's call into a search engine and being offered the best possible solution or answer from the system's knowledge base.

While it is almost impossible to make accurate predictions, as a single invention such as the smartphone has the potential to disrupt entire industries, we could well see automation remove the need for customers to interact directly with a contact centre completely.

As virtual assistants (VAs) such as Apple's Siri become increasingly prevalent and consumers become more comfortable with the idea of interacting with machines, the relationship that humans have with these machines will evolve. Rather than a consumer contacting a call centre, the VA takes responsibility for managing this process, either resolving the consumer's query itself or connecting with the contact centre via established computerised links. This effectively eliminates the need for a consumer to sit in a queue or work through frustrating IVR systems.

The potential of this automation technology is far-reaching, with benefits for both the customer and the contact centre. Not only does AI and machine learning allow for 24/7 service, but queuing for the next available agent becomes a thing of the past. Customers are also able to receive almost immediate resolution to their query, and in a language of their choice, providing a more personalised customer experience.

Naturally, the benefit to the business's bottom line is significant, considering agents typically make up 70% of a contact centre's expenditure. Shifting reliance to automated solutions means reduced premises requirements, reduced management overheads and reduced income losses due to shrinkage.

There is, however, an important aspect of automation that, if not given adequate attention, can become a considerable limitation on its success: automation and its component parts (virtual assistants, deep learning, machine learning and cognitive computing) are only as strong as the knowledge base on which they depend.

Forward-thinking contact centres should be looking at how they can prepare so that they are poised and ready to make the most of automation when the time is right. This means building FAQ and knowledge base engines that can assist human agents and be the basis on which all automation is built.

Beginning this process sooner rather than later allows contact centres to draw on their agents' expertise in ensuring the information contained in the knowledge base is accurate and comprehensive, so that it will be sophisticated enough to handle automation when implemented.

There is no doubt that automation needs knowledge, and having a successful contact centre of the future starts now.