Dynamic Business Logo
Home Button
Bookmark Button

The future of AI-powered chatbots in business: From niche processes to industry features

Less than a decade ago, the concept of an Artificial Intelligence (AI) powered agents answering customer questions in real time was still considered out of reach for businesses of all sizes. Today, these virtual agents, otherwise called chatbots, are very real, indeed with business adoption growing exponentially.

Last month NAB rolled out its ‘virtual banker,’ a virtual assistant using AI to answer more than 200 questions that were previously handled by call centres. And last October, Tennis Australia designed a new chatbot aimed at helping sell tickets to the Australian Open 2018 via social media. This drove 170 per cent more conversions than their traditional go-to-market model.

Through AI, chatbots will soon move from small-scale pre-programmed processes to being able to solve much broader industry issues, and deliver more personalised solutions to customers. As we wrap up 2017, now is a good time for business owners to look at the strategic investments they need to make today, to make the most of chatbots tomorrow.

Today, automated chatbots rely on computer coding and AI to solving specific, niche problems and requests like booking a restaurant, a taxi, or buying products online. To do that, they use pre-defined rules based on a decision tree framework. In practice, this means that they operate in areas of narrow expertise, based on scenarios of answers to common problems. They are not self-learning and self-sufficient operators…yet.

In the next three to five years, the growth of AI and other automation technologies will see chatbots expand these specific tasks to broader domains. Powered by machine learning, as well as deep learning – a technique which uses comprehensive networks to swiftly produce human-like conclusions – chatbots will be able to take advantage of vast amounts of data available across departments, as well as from outside the organisation’s walls. Combined with lots of processing power at internet scale, this will dramatically improve the quality of understanding and decision-making and tremendously deepen the range of capabilities they will be able to achieve.

Tomorrow’s chatbots will also hold sophisticated algorithms that will enable them to have a lot more context about each user and, in return, provide unique experiences for each person they talk to. They will be much more goal-oriented rather than taking customers through every step of the way. AI will bring that extra layer of intelligence, meaning they will learn about customers in real time. This will likely be used to create authentic, personalised relationships with a target audience.

Banks are already harnessing chatbots to help customers with simple tasks. In the future, they’ll be utilizing them for more complicated processes. Imagine an AI-powered banking bot that can provide broader personal finance advice such as how to save money and negotiate bills, not merely straightforward transactional inquiries. Such uses have the potential to revolutionise how certain industries manage their operations and deliver services to customers.

In an era of increased competition to deliver cut-through customer experience, businesses can’t afford to miss chatbots’ potential. Those that neglect to recognise it will inevitably lose competitive edge, not only through a loss of customers, but also through fading efficiencies and leading positions in their industries. The technologies necessary for developing intelligent chatbots are here: AI and machine learning. They are indispensable to the development of self-learning agents able to deliver personalised processes, as well as for creating integrated services to customers from scratch.

About the author

Mark Armstrong, Vice President and Managing Director of International Operations, APJ & EMEA, Progress. He previously contributed commentary to Let’s Talk… Disruption.

Related: Chatbots

What do you think?

    Be the first to comment

Add a new comment

Mark Armstrong

Mark Armstrong

View all posts