Top mistakes businesses make with AI in the contact centre, and how to avoid them


Johannesburg, 23 Nov 2018

There is a great deal of discussion in the marketplace regarding robots and artificial intelligence (AI) and their future role in the contact centre. Much of this is hype. A lot of people are talking hypothetically about what robots might do in a customer service context. Fewer companies are using a truly AI-driven approach to engage with customers today.

There is no one-size-fits-all answer here. Some organisations will continue to use human service as a key part of their value proposition and differentiation, but most are bringing in a growing element of AI and automation as they move to a more self-service-based approach.

As early as 2011, analyst Gartner was predicting that by 2020, customers will manage 85% of their relationship with the enterprise without interacting with a human. But, as they implement chatbots and other types of AI, there are a range of pitfalls businesses need to watch out for. Here, we outline some of the biggest and how businesses can best avoid them.

Failing to give bots a proper job description

Businesses are increasingly happy to bring robotic technology, including chatbots, into the contact centre to assist with their operations. There are significant potential benefits. Recent research by Accenture found CIOs and CTOs around the world believe conversational bots will play a critical role in the enterprise architecture of the future. Indeed, 60% of executives surveyed for the research said bots can "improve their organisation's ability to handle customer queries by networking with other bots", while 57% believed they can "deliver personalised attention to Web site visitors by being more conversational".

However, most bots fail to deliver in the first instance as most organisations don't link their acquisition of a bot to a defined business goal. Don't make the same mistake. After all, if you don't know exactly what you want the robot to do, how are you going to coach and develop it? Just like a human, a robot needs to have access to the latest relevant knowledge and information to do its best job. Otherwise, the value it offers will quickly fall away.

Also, think about how your new robot fits in with what we term the SPRO (strategy, processes, resources, organise) approach. From the strategy perspective, consider what you are trying to achieve, what business you are in, who are your customers, how best should you interact with them. From the processes point of view, what specific processes help you achieve these goals and interact with those most important to your business? You also need to consider what resources you need to best achieve your end objectives. Finally, you need to organise yourself to find the right way to deliver against the strategy, work out what you are trying to do and get your processes right.

Neglecting to integrate a robot properly within the organisation.

Many businesses implement robots and then neglect to integrate them properly with customer service operations. It is a mistake to think that robots can handle every interaction. They can't show empathy and they are poor at dealing with complexity. When mapping the customer journey and building a connected infrastructure to support that, you need to ensure there is a route to escalate any interaction seamlessly from a chatbot to a human, without losing context in real-time.

Forgetting the need to train

Just as with any human agent or interaction channel, you need to measure the robot's performance in delivering on business goals and then focus on making improvements. You wouldn't expect a new staff member to get up to speed on all your processes without internal training, so should you expect a robot to do the same? Treat your automation systems and your chatbots like human agents. Make sure you train and support them and regularly monitor their progress to drive continuous improvement.

Failing to use AI as an operational hub within your business

AI can help businesses by monitoring and understanding the customer interaction and using that understanding to provide better quality self-service. However, it can also help the business decide which human agent is best suited to handle every issue. It is not only about matching the skill set of an employee to the nature of the customer enquiry, but also matching the profile of an agent to the profile of a customer to help ensure the two 'get on'. That's important because whatever the issue, if people have a good rapport, the relationship tends to be more productive.

Neglecting the need to keep knowledge management up-to-date

In line with the 'garbage in, garbage out' principal, any robot is only ever as good as the knowledge at its disposal. You need to ensure that when a question is answered in the contact centre, that knowledge is captured and delivered into the knowledge management system (KMS), so customers, bots and human agents can feed off it.

If you imagine the automation effort within a contact centre as an iceberg, the small section at the top is effectively the artificial intelligence, while the larger section underneath is knowledge management. AI may have many benefits, but how can it be used to make decisions when it does not actually know anything? It can learn, but it needs relevant data to do that.

That is why it is so important to have processes and procedures in place that enable you to feed accurate data and intelligence into the KMS. Many businesses are too reliant on their employees as a source of knowledge. They have come to accept a situation where every day, much of the knowledge held within the business walks out the door and may possibly never come back.

Businesses need to find a better way of capturing that knowledge for good. Using knowledge management systems allows them to harness the intelligence in people's heads, so if those people move on, organisations are better prepared and more resilient in dealing with it.

Failing to guard against your bot 'going rogue'

Finally, while artificial intelligence and robotics technology is advancing all the time, businesses need to realise that their AI systems are not immune to being hacked. You need to guard against the possibility of your chatbot going rogue, for example. We are already seeing instances of hackers taking control of bots and feeding them with false information, potentially pushing customers in the wrong direction and to a fraudulent Web site, for example. You need to ensure you are putting the right security and threat protection technology in place to guard against this.

Looking ahead

There will continue to be hype around the use of AI in customer service operations. Robots will not be taking control of the contact centre any time soon; striking the right balance between automation and manned is a must for any organisation that wants to succeed. However, most businesses today are bringing in a greater level of automation into their customer-facing operations and many are making growing use of chatbots and other AI-driven systems to give customers what they need. That's a big responsibility for any business, and as they deliver on it, organisations need to avoid the common snags and stumbling blocks that many fall prey to.

Following the above tips should help them stay trouble-free as they plot their route to a more automated future.

If you would like to see how our customer, Fujitsu, went about installing and running chatbots within the contact centre, download our Webinar on: "Chat Bots and AI in the Contact Centre".

Share

Editorial contacts

Patricia Merchan
Enghouse Interactive
Patricia.Merchan@enghouse.com