So where is AI at now in the contact centre world, and what can we expect in the future? Is it possible that one day robots and other forms of AI will totally take over the industry, leaving swathes of contact centre employees out in the cold?
Rise of the internet chatbot
To the first question, AI is largely still in its experimental stage in Australia. But the results to date have been promising, particularly when it comes to text-based chatbots (or ‘virtual assistants’, as they’re also called). For customers preferring to have their questions answered via a website than by calling a contact centre, chatbots can provide excellent service.
When a chatbot pops up to a website, it comes with Natural Language Understanding (NLU) capabilities that allow it grasp the intent of what’s being asked. In other words, it mimics how a knowledgeable human would respond to the enquiry. A good chatbot is also able to seamlessly transfer the customer to a live agent when the conversation gets complicated.
An example of a successful chatbot in action is Alex, the virtual assistant for the Australian Taxation Office. Alex uses NLU, advanced resolution technologies and conversational dialogue to answer tax-related questions from website visitors. Launched in March 2016, Alex has answered over a million enquiries with a resolution rate of 80%, which is significantly higher than the industry benchmark of 65%. This type of technology has also gained traction in the telecommunications, banking, retail and financial services sectors.
According to recent research by conversational commerce company Flamingo Customer Experience, 77% of Australian customers (from 528 surveyed) are comfortable with the concept of using a chatbot to address their enquiries. Of the 84 Australian businesses surveyed, 73% believe chatbots are relevant to their business. With a solid track record already laid down, and with the technology improving all the time, internet-based chatbots are here to stay.
Live agent or robot – what’s best?
While chatbots can be good at providing simple answers to predictable, uncomplicated questions, the stakes get higher when organisations leave it to a robot to address complex enquiries. Companies don’t want to risk losing a long-time customer because of a poor Chabot experience. They also don’t want to fail to convert a prospect into a customer, because of under-developed voice automation technology.
This is one of the reasons why Australian organisations are treading carefully when it comes to voice-based AI. As things presently stand, robots simply aren’t good at certain things. Through machine learning, they can be effective in understanding what we said, but not necessarily what we meant. For example, AI technology is typically poor at detecting sarcasm – a response of ‘yeah, great’ from the customer may not mean what the robot thinks it means. An empathetic understanding of a customer’s emotional state is also not something a bot can easily pick up on. So a customer becoming increasingly exasperated may not be dealt with in the same way as they would if there was a human on the other end of the line. Also, a robot can only provide answers from which it is able to draw from its database of knowledge – a potential shortcoming when a customer makes a ‘first-of- its-kind’ type of enquiry. While machine intelligence operates within certain pre-programmed parameters, human beings do not. We’re simply not that predictable.
For contact centre operators, a key challenge lies in knowing the point at which a customer machine conversation is failing and needs to be taken over by a live agent. If a live person is too slow to get involved, the customer could be lost.
So until major further advancements are made in AI, its core strength will remain its ability to automate simple processes where human skills and knowledge are not required. This is good news for the contact centre workforce because there’s no way automation is going to replace humans any time in the foreseeable future. Rather than replace live agents, technology can be expected to complement and enhance both the customer experience and the agent experience. That all-important need for the human touch, where qualities such as compassion, an instinct for social cues and a grasp of the unique needs of a particular individual customer, will continue to be vital for organisations dependent on attracting, serving and retaining customers.
For contact centre representatives, automation will free them from the tedium of continually responding to the same rudimentary queries over and over again. Their jobs will become more stimulating and challenging, allowing more opportunities to apply critical thinking and creativity in their problem-solving.
Is AI a win-win for customers and providers?
For customers who want good customer service and for contact centre service providers, who want to deliver it, the future looks bright. For the most part, Australians are comfortable having their basic enquiries dealt with through automation. Not speaking to a human being is generally acceptable as long as it provides the desired result. In this regard, AI has plenty of runs on the board, and will only get better as its machine learning technology improves.
For government and corporates, AI has delivered some successes in reducing operational costs, while maintaining solid customer satisfaction rates. With the high cost of hiring and training contact centre staff, the reliable and efficient automation of basic functions has predictably had a positive bottom line impact.
But there is still some distance to travel along the AI path. Consistently outstanding customer engagement from an intelligent, insightful and empathetic robot is the holy grail. But, in all likelihood, it’s destined to remain elusive, despite the impressive pace of AI innovation.
The contact centre of the future is far more likely to be one that combines human interaction with automated processes to deliver the best in customer service.