How will AI influence what our customers expect from us?

Artificial intelligence (AI), once relegated to science fiction, is now a common part of our daily realities. For the average person, encounters with AI will often come through interactions with contact centres, whether they even realise it or not.

With the increased efficiencies created by AI, expectations concerning customer experience are also changing. So how is AI actually influencing what our customers expect from us?

Phil Jones, Manager of AI Solution Development at Stellar, believes there are four main areas where AI is impacting customer experience (CX).

Firstly, Phil says, “We’re seeing trends around personalisation being a bigger part of customer experience. If companies already have information they’ve logged from previous purchases it can be used to predict decisions. That’s where AI can help.”

“Secondly, there’s a lot now in terms of internal knowledge bases in companies – that’s people being able to use self-service or virtual agents that have picked up specific questions people have asked over time. That sort of thing is going to become more important.”

Phil admits we’re going to see more and more examples of artificial intelligence at the heart of customer interaction. Amazon’s Alexa – Echo speaker – is a current example of how AI is infiltrating the home and influencing the knowledge inputs of the average person.

“In some cases, technology like Alexa is driving self-service conversation but it’s also driving conversations through to contact centres,” he explains. “You might just ask Alexa to call someone and you’ll be put right through.”

Phil points out that the inherent distrust people still have about the reliability of decisions made by AI. This distrust means humans are unlikely to be out of the loop, even as AI develops in its sophistication. 

“There’s still human apprehension,” confirms Phil. “One of the things we’re working on is the computer modelling behind responses that agents might provide back to customers. It is becoming an increasing focus, having people go in and ‘train’ those models – human-in-the-loop (HITL) – and adjust the responses they give to make them more accurate. That’s the categorising of the data so the responses are sorted, which means it takes out – for example – biases and inaccuracies that are coming through AI.”

Erica Johnson, GM Digital AI at Stellar, says customers have “a level of tolerance” for AI, which she considers to be acceptance, rather than expectation.

“Customers would be happy to deal with other humans for simpler issues as well but we’re increasingly taking that option away from them,” she says. “Part of the reason for that is because we believe the results from AI-powered solutions will be better. You remove some of the human error factor. You get a higher level of predictability of outcomes.”

“Through human-in-the-loop, we’ve got a way of designing, and controlling what kind of accuracy and outcome a customer receives from an artificial assistant. That should give us greater consistency compared to having an agent remember the training they had some time ago.”

Erica says the market is at an interesting point where AI has opened up to tech start-ups due to intelligence and capability being made publicly available through tools and algorithms. Consequently, a number of fragmented players have emerged in the marketplace, tackling really specific business problems rather than a holistic solution. 

As Erica observes, “You’ll have a company who might be dealing with the compliance side of quality monitoring, and someone else will be dealing with the speech analytics side of monitoring and customer sentiment, and someone else dealing with processing of the customer enquiry and so on. There are all these examples of start-ups that have found a specific business problem and they’re going really deep on solving it.”

“These small companies are solving one kind of problem but, when I solve that problem, I’ve got another problem that runs alongside it, and I need a different partner for that,” continues Erica. “So, short of being one of the big Amazons or Googles, businesses in the small or medium space who want to be able to deliver better customer outcomes, it’s almost overwhelming knowing which problem you start with and who you partner with.”

According to Erica, customer attitudes towards AI differ from those of businesses. As customers, we have a reasonable understanding of the level of maturity of AI, and we’re willing to try something again further down the track if we don’t initially get a satisfactory result. However, business managers are the harshest critics of AI and expect perfection from the onset.

“There’s a real nervousness when you release it to market that it has to be perfect,” says Erica. “We don’t want to put it out there as an incomplete solution. We are fearful of what people will think if we put something out there that is not able to meet all their expectations, whereas that seems to be in contrast to the way our customers think about it.”

“There are different expectations across the AI space. As designers, we should be thinking about it more like – in the contact centre space – how we think about new agents. When you hire someone new to your business, you don’t expect them to hit the ground running on day one – you expect them to get better over time. We should have that same relationship with AI.

When it comes down to it, all people want from their customer experience are great outcomes – “but it’s not that they expect us to deliver them with AI,” concludes Erica. “It’s more that they will accept AI if the outcome is what they’re looking for.”

About The Author

Stefanie Cutrera
Stefanie oversees the internal communications strategy and execution within our business, managing all communication platforms and setting the gold standard in getting information disseminated to the right people.

When she’s not devising clever new ways of opening up communication lines at Stellar, Stefanie loves to travel.

Want to learn more about Stellar?

Get in touch with us today.