From Westworld to Westfield: AI is transforming the retail experience
AI has fascinated humankind for decades, as we both dreamt and feared what a programmed intelligence could be capable of. For many years this speculation was strictly the stuff of science fiction, but now technological advancements have taken AI off the page and into our everyday lives.
A wide range of businesses are actively experimenting with how both AI and other exciting technologies such as augmented reality (AR) can change the way they communicate with and sell to customers.
But it’s the retail sector that’s in the vanguard of this revolution.
AI has the potential to transform how we shop both in store and digitally.
Already, online AI programming is making recommendations, based on items a customer has purchased or looked at. As this technology develops customers will be offered more nuanced recommendations ̶ for example someone looking at a top could be given suggestions for a belt, bag or scarf that complement it.
shoppers will be offered tailored style recommendations from an algorithm
In the near future, shoppers will be offered tailored style recommendations from an algorithm that is capable of learning over time how people have combined items.
These suggestions could also be offered to a customer in-store, where digital mirrors will show items that would look good with what you’re already trying on. Or, using augmented reality a customer could see what other products would look without actually trying them on.
AI is changing how we shop beyond fashion; it’s altering regular purchases such as the weekly grocery shop, too. The more we trust in online retailers and share our data with them, the more they can tailor the timing, content and context of their recommendations.
AI will enable retailers to predict when we will need items
By knowing our buying habits, AI will enable retailers to predict when we will need items ̶ for example, if we normally buy bathroom cleaner every four months, after three and a half months your supermarket could ask if you want to add some cleaner to your trolley.
Algorithms are constantly learning, becoming smarter and making better recommendations, and the more we interact with them the more astute they will become.
See it, find it, buy it
It’s a common experience to see someone wearing something you love, but you don’t know where they bought it.
Currently you have two options. One is to march up to the stranger and ask to know where they bought said item, or alternatively trawl the shops in the hope of finding it.
visual search aims to make it much easier to find what you’re looking for
But this is set to change. Visual search, currently being trialled by Asos, aims to make it much easier to find what you’re looking for. It works by comparing the pixels in imagery to identify and return results that are similar. You take a smartphone photograph of the item you want and the algorithm will search for it online.
Instead of searching by key words that could produce thousands of results, visual search will be able to narrow the results down to something more specific. Searching by image rather than by text will make it easier and faster than ever before for consumers to find what they want.
Talk to the machine
A number of retailers are experimenting with chatbots as their first dip of the toe into AI waters. A chatbot is a computer program that can hold conversations either by text or audio.
by 2020 the average person will have more conversations with bots than with their spouse
Capable of answering basic queries, retailers are increasingly using them to help with customer service. Research and advisory company Gartner have suggested that by 2020 the average person will have more conversations with bots than with their spouse.
ShopDirect recently announced it was investing in chatbots, as it felt customers wanted an instant interaction and not to have to wait for an email response. They allow customers to talk to a business online, getting a response to a question or query immediately.
At the moment responses are limited, but the more interactions chatbots have the greater their repertoire will become.
A system is only as good as the data you put into it. We need to take great care that the data we use to build AI programs for retail is representative of the results we hope to achieve.
a system is only as good as the data you put into it
Last year Microsoft AI chatbot Tay had to be taken off Twitter after posting a slew of racist tweets. Tay wasn’t programmed to be racist but online trolls exploited the fact the chatbot didn’t understand what it was saying and taught it to say highly offensive things.
There is a duty of care in the training of any AI; in some ways it’s like a child, learning from a simple set of rules. Beyond such conscious bias, we also need to be mindful of unconscious bias.
Historically the people developing both speech and facial recognition systems have been predominately white men, and one consequence of this is that facial recognition software is better at identifying Caucasian faces and voice recognition software is better with male voices. If retailers plan on using these technologies to enhance their customer’s experience these biases must be addressed.
We’re still in the early years of AI in retail, and while it’s being adopted quickly by some brands, it has yet to be been taken up broadly by the industry.
letting an algorithm decide where to place your adverts and what products to recommend
Retailers see the potential but there’s uncertainty in how to how to make it work for them and their customers. Wide-scale AI adoption requires a step change in both business and customer thinking -̶ we need to trust a machine to make certain decisions which up until now have always been made by a human.
For retailers, that could be letting an algorithm decide where to place your adverts and what products to recommend, or when to discount a product and by how much.
For customers it’s a willingness to accept fashion advice from a computer code that doesn’t wear clothes, and being willing to seek help from a chatbot with a limited range of responses.
Technology is a useful and powerful tool. Recent advances could forever change the way we shop, making it quicker, easier and more personalised to buy everything from a designer handbag to your favourite bottle of wine.
While algorithms excel at pattern recognition, humans have the capacity to be creative, demonstrate empathy and understand their fellow beings in a way that is ̶ so far ̶ inconceivable for a machine. In the age of automation and artificial intelligence, the need for human understanding has never been greater.
- » How to build the best martech stack for data-driven marketing: A guide
- » Why empathy is the most underrated weapon in the marketer’s arsenal today
- » How artificial intelligence drives genuine ROI from real customer feedback
- » Five key tips to drive value from content and martech experimentation
- » Building lasting customer loyalty through personalisation: Why CXM is your new best friend