Paul Mahon_ The SBPost

Paul Mahon, our Technology Consultant, was featured in this month’s Business Post Connected – Customer Experience.

AI has been seen as the future of customer service, but businesses should not put unreasonable expectations on what is still a young technology, writes Quinton O’Reilly from the Business Post.

When you give it simple outcomes, AI can be a huge benefit, but if you over-extend it, things can fall apart, said Paul Mahon, senior technology consultant for Singlepoint. It’s a tool that can be used for specific situations, yet it’s nowhere near the point where it can deal with unknown variables properly. “To use the self-service checkout example, the challenge becomes that as soon as you diverge from the well-de-fined path – it doesn’t recognise the bar-code or the object isn’t heavy enough and it doesn’t register – it won’t progress until you do that,” he said. “Those are those edge cases where these things fall down.” That’s the issue all technology faces, to be so reliable that the frustrating ex-periences are as close to non-existent as possible. Rightly or wrongly, the times it goes wrong are the moments customers will remember.

To help mitigate that, it’s about defining expectations. Part of why self-service checkouts are frustrating is because they promise an experience that’s faster than a person checking out your items. Instead, we remember the times it frustrated us. AI can avoid that pitfall if it’s applied correctly. If it gets the right kind of information, it can either solve a problem directly or give a human agent enough data to solve the problem. It could reach the point of adapting accordingly, but for now the focus should be on addressing those low-level concerns.

“When you’re trying to provide an exemplary customer experience, what sticks in your mind is when it doesn’t work rather than when it does,” said Mahon. “What companies are struggling with now is how they enrich and augment the experience when using a chatbot as their starting point.” In the case of chatbots, more companies are being upfront with customers about who they’re talking to, which can put them at ease. If they know they’re interacting with a bot, they can adjust their expectations accordingly. “In the early days, the thought process in management was that you shouldn’t be telling the customer that they’re not talking to a user,” said Mahon. “Now the concept has evolved to the extent that being upfront is more commonplace. The customer is a little bit on guard, but probably ends up more accommodating as a result. If the conversation flows, they’re impressed, but if it doesn’t you can get a human to handle the problem.”

If you would like to speak with Paul Mahon or any of our team, please contact

Upcoming Events

DAPx Webinar  On-Demand

Watch here!