Now that frenetic Black Friday and crazy Cyber Monday are behind us, I’ve decided to complete my Christmas shopping. And this year, I’ll be doing it with the help of artificial intelligence (AI). With the tech giants, such as Google, Microsoft, Apple, and IBM, increasing their investments in AI and news of Alibaba’s $15bn investment in it, I wanted to find out what the customer experience is really like. And, most importantly, does it live up to all the hype?

What is artificial intelligence?

So, what is AI? The term AI was first used back in the 50s and was based on the principle that ‘every aspect of learning or any other features of intelligence can in principle be so precisely described that a machine can be made to simulate it.’ Essentially, it’s about machines making calculations and decisions by mimicking human decision-making processes.

Machine learning, which is sometimes confused with artificial intelligence, is actually a current application of AI where machines are given data, and are then directed to make predictions on new data, based on learnings from previous data.

Today there is a lot of buzz about AI because of the mountain of data available through the internet, and the increasing power of computers.

Instead of coding a set thinking process, engineers are coding machines using neural networks so they can think and understand the same way as humans, but with more speed, accuracy, (memory) capacity and with no bias.

Combine this with Natural Language Programming (NLP) and you get a machine that can not only understand but also communicate like humans. A bit too much sci-fi? That’s in a way what it is, but not as we imagined it…

Let’s take a look at how AI is affecting the shopping experience for customers, by seeing if it can help me find the right presents on my Christmas shopping list:

  1. A new ski jacket for my husband
  2. Something for my mum
  3. A gift for my sister

Virtual personal shopper

A virtual personal shopper provides a personalised service by learning about what customers like, how they behave and providing tailored recommendations to customers.

For a new ski jacket for my husband, I’ve decided to try Northface XPS, the Northface Fluid Expert Personal Shopper, which is powered by IBM Watson. It offers recommendations based on a series of answers to questions about where, when, for whom and what purpose the jacket will be used. The questions are simple, easy and straightforward to answer, as if you are talking to a sales person – e.g. Where are you going? When? What will the jacket be used for? And then it makes a recommendation based on the destination, average temperature and purpose.

Verdict: Easy, simple and intuitive. Gift number one, ticked.
Robo rating: 4/5 

Conversational commerce

Conversational commerce allows customers to interact with an AI robot and ask questions, receive personalised recommendations, and place orders. This can be done through voice, like Amazon Echo (Alexa) or Google Home, or via instant messaging, such as Facebook Messenger or WhatsApp. It can also include human intervention when there are exceptions.

WeChat is the ultimate in chat commerce. Currently in China around 600m people are using Conversational commerce via WeChat every month. Hailing taxis, booking doctor’s appointments, and paying utility bills are all performed via chatbots.


Chatbots are essentially another type of virtual personal assistant or shopper. A chatbot lets you buy a product through a chat function. For those who are not fans of social media and instant messaging, get used to it! These are here to stay, and in fact, are predicted to grow. According to The End of an Era for Apps they will soon overtake apps and a study predicts that by 2020, 80% of customer interactions will be handled by artificial intelligence.

As I needed inspiration for a gift for my mum I decided to try out the eBay ShopBot. The chatbot asked me a series of questions that were totally unrelated to the practicalities of gift shopping (gender, budget, interest), and were more about the lifestyle preferences of the recipient – e.g. What they like to do at the end of a long day? What type of hotel do they like to stay in when they travel?

Following the set of questions, which are accompanied by GIFs to make the experience more light-hearted, the ShopBot describes the persona based on my answers, and offers a gift selection based on that persona.

Verdict: I think chatbots (at least those for shopping assistance) have a lot of learning to do before they can be mainstream. They are useful for when you know what you want and just need the AI to search for it. However, when you need inspiration, it’s not the best experience – yet. I don’t doubt that through machine learning, they could get it right, eventually. But for now, I’d rather walk through the physical stores to get inspired…
Robo rating: 1/5


Visual search

Visual search may help with the inspiration challenge. Visual search works when a user uploads a photograph of an item they like or an image from other sources such as Instagram or Pinterest, and through image recognition technology, the AI recommends similar products.

I was looking for a gift for my sister and was inspired by a pair of shoes I saw on someone on the Tube, and photographed them. I tried out the visual search function on the ASOS app, and uploaded the photo. It then recommended similar items on the ASOS website.

Verdict: It works best when a third party provider does the visual recognition and suggests items across many retailers, to search through the widest possible options. However, there is something to be said about curation that a trusted brand provides. And at ASOS, it was a pretty slick experience, and quite accurate. Yes, I had to have the ASOS app, but apart from that, it was pretty easy – open the app, tap the camera icon, tap Photo Library select photo, and click ‘choose’. Gift number 3: Success!
Robo rating: 3.5/5

So what?

What’s clear to see is that the consumer-facing benefits of AI are still in their nascent stage, with only a handful of companies testing and learning different ways to personalise the shopping experience.

However, since it is called machine learning, that is precisely what will happen. The technology will only continue to improve its intelligence over time – quite quickly in fact. Add to that things like sentiment analysis and facial emotion recognition, recommendations could be even more relevant and accurate. And the experience could be brought to the shop floor. Although frankly, that could be a bit creepy…

I do foresee a world however, where I have my own virtual personal shopper for all my needs (including my weekly grocery shop) at my fingertips. And where inspiration from my surroundings can convert easily into purchases or experiences, with just a few clicks. Now that’s a different way to shop.

Retailers, are you prepared? Are you are thinking about how artificial intelligence will change the way people shop, and how it will consequently shape your business going forward? AI will impact every aspect of it – how you engage with customers, how you operate, how you make decisions, and the type of skill sets you will need. You need to act now.