Back to articles

How will AI & deep learning impact the customer experience?

29th June 2017
By Steve Kilpatrick
Co-Founder & Director
Artificial Intelligence & Machine Learning

Personalisation is one key element to customer experience that cannot be emphasised enough. The tendency, however, is for marketers to begin to see personalisation simply in terms of the advantage and the value that brands can acquire from deep customer data.

What we must not forget to ask ourselves is what personalisation means for the consumer themselves. And this is a question that should remain at the forefront of our minds as we progress into an age where deep learning and artificial intelligence becomes ever-more ingrained in the customer experience.

Shared value exchange is at the heart of customer experience, particularly when it comes to data. What this means is that, in order for a customer to be willing to share their information with a brand, they must feel that they are getting something of value in return.

As more customer experience becomes mediated through AI, chatbots, virtual assistants, and machine learning algorithms in general, this sense of shared value becomes more important.

There is a risk of disengagement if customers feel that their custom is not of value, and this can be a real danger when the job of interacting with customers is placed in the hands of, not a human, but a robot. In short, there are plenty of ways that AI and deep learning could impact the customer experience. The trick is to make that impact seamless.

I recently saw discussion from Warwick PhD student, Henry Powell at the RE:WORK deep learning conference.

Powell’s research is on the philosophy of cognitive science, neuropsychology, and robotics. He looks at the way in which low-level dynamic and computational control of motor action can affect human-human and human-robot interactions. Also, for social AIso, he looks at how human-like commitments can be engineered into social-robotics platforms to engender beneficial human-robot interactions in medical, commercial, and social spheres. An example would be:

On a messenger app, you wouldn’t expect a person to reply instantly with a lengthy answer to your question. You would expect them to think and then type. If a person senses that they are talking with a robot, they tend to become disengaged. He focuses on trying to make them more human feeling, by assessing factors such as the optimal time for the AI to reply to a question, and so on.

Research like this is what will drive this balance of shared value between brands and consumers when using AI software in customer-facing positions. However, text-based chatbots in messenger apps are just the tip of the iceberg for what’s to come.

So, as we reach a more AI-powered future, where will brands be connecting with customers to gather and analyse this data?

It is anticipated that screen time will decline as we reach a point at which interactions will no longer require a keyboard. Instead, gesture, voice, and interactions through virtual and augmented reality will be the principal interfaces with which we interact with brands. This evolution will occur in tandem with the growth of the Internet of Things.

At a recent DMA (Direct Marketing Association) event, Jeremy Waite, Evangelist at IBM Watson, boldly predicted that by 2019, one million new devices will be coming online every single hour. These devices are not desktops, laptops, tablets, or even the currently omnipresent smartphone. The devices Waite is talking about are the connected devices in our homes, our cars, and our offices, that will be in constant communication with one another.

In the smart, connected future, brands will not be competing to place their products in front of us – they’ll be placing them all around us. How does that work? Well, through artificial intelligence, of course.

The Seamless Ambient Marketing Era

Say, one day, you take delivery of your grocery shopping to your home. It’s an order you’ve placed by adding stuff to your shopping list using your virtual home assistant (Alexa, or whatever her alternative/successor may be), and you’ve included all your usual favourites. But as you’re putting the shopping away, you notice there’s something you hadn’t ordered.

“Alexa,” you say. “What is this chocolate bar? I didn’t order this!”

“Well, Dave,” Alexa replies. “I notice that you like raisins and hazelnuts, and this new bar has both of those. Plus it’s Fairtrade, which I know you prefer.”

So, you bite into your chocolate bar. It’s good. Your friend, Bob, would like this.

“Alexa, send one of these to Bob’s shopping basket,” you say.

“Of course,” Alexa replies. “You can also send it to two more friends for free within the next 30 minutes.”

You send Bob his free chocolate bar. You also send it to Jane and Paul. Now all three of them get to try the chocolate bar, and offer it to three of their friends for free. Everyone has a great customer experience with the new chocolate bar, and next time, they may just order it for themselves. This time, not for free.

This is a very basic example, but you get the picture.

Another might be that you’ve got a hiking trip booked into your diary in a few weeks’ time. Your bedroom mirror picks up on this and assesses the weather and conditions in your destination at the time of your trip. It also assesses your wardrobe, and finds that you are without appropriate walking boots. So, as you’re getting ready in the morning, your mirror gently mentions that you might want to order some walking boots for your trip. “These ones from Timberland are on offer,” the mirror says. “Shall I order you a pair in your size?”

Artificial intelligence is hard at work here, working out what products are right for you and when. Deep learning allows it to remember your size, your preferences, and when you are likely to be most receptive to interaction. This AI is not intrusive, it is not pushy. It helps you make the best decisions for you, in a way that is both enjoyable and useful. It delivers the ultimate customer experience, and you didn’t give its artificiality one single thought.

The Journey To Ambience

At this early stage of AI adoption, consumers veer between curiosity and wariness. Speaking to a chatbot has novelty value, but the sense that their interactions with a brand are being controlled by deep learning algorithms, to put it bluntly, gives them the creeps.

As such, there is a balance to be reached, between offering customers innovative features and experiences, and not coming across as intrusive. This boils down to delivering the right content at the right time, when customers are more responsive and appreciative of the brand offering, rather than simply in real time, whenever and wherever that may be.

Where this balance is achieved, and the shared value exchange between customer experience and data is in equilibrium, the most positive use of artificial intelligence and related technologies is found.

Our current and developing machine learning tools are allowing us to gather and process more data than ever, and crucially, to process it with unprecedented accuracy. In an ironic twist, the technology that will allow brands to target consumers more sensitively and engagingly would put customers off if they were to see it in action.

Human intervention and control must, therefore, support AI and its adoption in order to build a strong foundation for great customer experience. AI can analyse data on all sorts of factors, including human behaviour and emotions, but it is the real human touch that must pull the strings in order to turn that data into meaningful customer experience. The Wizard of Oz is nothing without the man behind the curtain.

I’d love to hear what you think…

Share Article

Get our latest articles and insight straight to your inbox

Hiring data professionals?

We engage exceptional humans for companies looking to unlock the potential of their data

Upload your CV
One of our consultants will contact you
to find out more about you