AI is becoming entrenched in every single aspect of the retail user experience, from online shopping to customer services and now even bricks-and-mortar stores (as was shockingly revealed by Amazon late last year).
From that very first Google search to clutching that shiny new product in your hands, AI is with you every step of the way. Let’s follow the hypothetical journey of Bob looking for some sun glasses.
Bob uses the IoT as his PA
The Internet of Things (IoT) has blurred the lines between operational technology (stuff) and informational technology (in this case, the internet) and is set to bring on the fourth industrial revolution. The IoT controlling your heating and alerting you about a burst pipe is just the start – in fact, it’s been projected that by the year 2020 there will be 24 Billion IoT devices installed.
In this case, Bob’s in the kitchen and is chatting to his AI assistant Alexa via Amazon’s Echo. Bob asks her what the weather’s looking like over the next week and she tells him it’s looking like a scorcher. Equipped with this information, Bob decides he needs some new sunglasses.
AI in search engines
So Bob goes on Google and starts looking for sunglasses. Bob loves his colloquialisms, so instead of looking for ‘designer sunglasses’, he types ‘designer shades’ into the search bar. This is where the AI journey starts.
While we all know how notoriously secretive Google are about the factors its algorithm looks for when ranking a website, one factor Google has been vocal about is RankBrain, the search engine’s machine-learning AI system that was rolled out this year.
Using the above search term as an example, through machine learning RankBrain has deciphered that in this context, ‘shades’ means sunglasses and has intelligently produced relevant results.
Another factor that RankBrain has considered as it looks for relevant websites is their respective structures in comparison to the industry leaders in that vertical. So just for fun, let’s say my sunglasses website (waleedssunglasses.com) is considered the industry’s leader and has historically provided the best user experience. RankBrain will value websites with a similar structure to mine, because if my website served searchers well, surely similar websites will do the same.
Bob arrives at waleedssunglasses.com and is greeted by a chatbot
It’s clear from the fast development and growth of chatbots in 2016 that they are going to play a major role in the future of UX. Through a messaging medium such as Facebook messenger, customers can type a query in natural language and get a response from a bot rather than a human. This is an exciting prospect for small and large businesses alike. With improvements in the technology, chatbots will offer a scalable, cost-effective and ever-present customer service solution.
In Bob’s case, it’s 2am (Bob’s a night owl) and he really doesn’t know what he wants. Luckily, my chatbot is on hand to offer assistance.
By asking Bob a few simple questions (what his budget is, what occasion the sunglasses are for, what styles he prefers etc) my chatbot is able to offer Bob exactly what he is looking for.
But Bob finds it difficult to make decisions, so wants to see a few more similar styles
Big Data, computer vision and artificial intelligence are being used to develop solutions that not only make the user experience better but save companies time and money.
For instance, just like John Lewis, eBay and Virgin Money, I have decided to use the visual search solution created by tech company Cortexica to help my customers find similar items and to help my business more easily categorise stock.
Cortexica mimics how the brain’s visual cortex perceives objects and images, and combines that with its identification ability from machine-based learning. In other words, their solution can look at an image of a product, identify specific features within it and then use AI to categorize it accordingly. Cortexica’s technology has 95% accuracy, the same as a human being, but is a whopping ten times faster.
Thanks to Cortexica, my website has given Bob a nice selection of matching styles to choose from. Bob has had his needs accurately served, and happily makes his purchase.
The future of AI always seems to go ominously hand-in-hand with the end of the world (but that’s an article for another time). However, if using AI in retail is giving customers a better service and helping them find exactly what they want, can it really be seen as a bad thing?