AI: When Over-Engineering Fails

Success in developing artificial intelligence (AI) would be the greatest event in human history. Stephen Hawking, the famous English physicist, once said so, and it seems that success is becoming more and more a part of humanity's future. The benefits of AI are growing rapidly, and it is being applied in many fields, most notably in financial technology (Fintech). However, while the benefits of AI are obvious and will be even more obvious in the future, there is one detail that I would like to highlight at this point. This is the application of AI in client support because some things show that the growing use of AI is not yet superior to traditional, live communication, face-to-face consultations and etc. I want to share some nuances as to why this is so.

 First of all, the growth of AI itself as an important technology is well illustrated by a recent study carried out by KPMG together with research platform Statista in 2021. 35% of respondents to the survey said that the adoption of AI is in full swing, with 49% citing AI as increasingly functional in their company's operations and 13% as just gaining momentum. And just 3% said they do not use AI at all. 

 Another survey of a similar type showed where AI is being used in the financial industry. Of the 1,000 or so respondents, 58% said they use AI for fraud detection, 41% for financial process analysis and 33% for cyber security. 31% used it for more detailed customer requirements and analysis.

 However, the most important element that demonstrates the comprehensive importance of AI as a technology and the extent to which AI will be relevant in the future is the revenue generated by the AI software market worldwide. In 2018, revenues were estimated at USD 10 billion, while it is now predicted that they could exceed USD 400 billion in 2022. Interestingly, just a couple of years ago, it was only cautiously predicted that in 2022 it would surpass USD 100 billion. 

Such statistics are more than a clear demonstration of how relevant AI as a technology is now and will be even more so in the future. It is simply impossible not to admire this. However, even as we enjoy such rapid advances in AI, we are encouraged to take our time by the clients of financial institutions, who clearly show what is currently most important to them in customer service and what is not yet fully covered by AI.

What do customers say? Things that have always been commonplace, such as live communication and consultants, are still extremely important details even in the context of the rapid growth of AI. It's also a kind of message that it's good to improve, but you don't have to be "over-engineered" because live communication these days is not about stopping in time and saving money. On the contrary, in-depth and warm live communication and consultation are still a great investment in increased customer satisfaction.

Just recently, my colleagues and I had the opportunity to conduct a survey to find out why customers choose one or another financial institution in general. And that survey clearly showed the importance to the customer of the close and human contact provided by live customer service specialists.

As many as 80% of respondents ranked a greater focus on customer service - understanding, flexibility, courtesy, etc. - as a key focus when choosing a financial partner. The same percentage of respondents said that when dealing with a flexible, close and understanding advisor, it is much easier to listen to needs and solve problems more efficiently.

Do these survey results matter? Very much so. For example, according to a Harvard Business Review study, brands that are able to maintain a high level of customer satisfaction are able to grow their revenues 2.5 times faster.

Yes, when it comes to customer service, AI with features like chatbots saves a lot of resources. Yes, they are available around the clock, they are able to gather more information and etc. But this is still at a fairly early stage of technological development. Perhaps that is why people sometimes even joke that talking to a fictitious consultant is like talking to a small child because the same answer applies to most questions. 

The still low level of trust in AI-generated features in customer service is confirmed by a study of more than 1 000 people in 2021 by a company that develops software to improve customer experience worldwide. 

As many as 60% of those surveyed said they would be willing to wait in a queue for a while if they could talk to a live consultant immediately afterwards. The same percentage said they would have a positive attitude towards the availability of AI-driven features if they led to the possibility of contacting a live consultant. Most importantly, more than 50% of respondents said they were generally dissatisfied with interactions with robots because, for example, the chatbot function did not understand the question and was unable to offer a suitable solution. In addition, more than half of the respondents said they wanted a clear understanding of when they were communicating with a robot and when they were communicating with a live person.

I believe that greater trust in AI is only a matter of time. AI-based solutions and their importance are on the rise so much that sooner or later, they will have such a high quality that people's expectations will be fully met too. But current reality, based on customer surveys and statistics shows that AI in customer service has not yet evolved to the point where it can meet absolutely all needs.

What does all these things say? When we enjoy the tremendous growth of AI, which is particularly evident in the Fintech industry we need to keep in mind not only the resulting speed, flexibility and analytics but also the simplest things such as understanding, courtesy and warm communication. And it is the latter that people still tend to notice more clearly in live interactions, but not through AI technology. 

The moment when AI is fully capable of operating in customer service and completely replacing the functions performed by live advisors will surely come. But as many people like to say - everything is good if it comes at the right time. So there's no necessary to be fast. We just have to clearly see the situation and just to wait.