AI disruption
The Evolution Of Artificial Intelligence In eCommerce

There is no denying that advancement in technology is one of the key driving forces behind the 10% year on year growth in eCommerce sales which is expected to reach $706 billion annually by 2022, according to Forrester Research.


There is also no denying that Artificial Intelligence (AI) and machine learning are defining factors in this extraordinary growth.


From predicting user behaviour to intelligent product suggestions, the speed it takes for a customer to pass through a digital shop is increasing all the time and not because they dislike it and want to leave as quickly as possible, but because merchants are able to put the right products in front of the right customers eyes more accurately and with greater speed.


artificial intelligence digital image


When was Artificial Intelligence created?


AI has been used in eCommerce for many years, one of the most obvious examples being personalisation through targeted products, which has infuriated many for the last decade or so.


It is now rare to land on an eCommerce site that doesn’t implement AI somewhere on their site and even off their site in that retailers and brands are are able to personalise the experience of the search results pages or the category listing pages based off the visitor behaviour from their website.


AI technology is increasingly set to replace marketers and merchandisers in selecting which adverts to place where and at which time across the Internet.


Predictive analytics is an extremely important field of AI and increasingly it’s becoming more accessible to smaller companies. It gives the opportunity to provide deeper insights into what consumers want and need and suggest to marketers and advertisers what the next action should be. So if you are able to analyse all of the activity on a website and serve this to a merchandiser and say which customers are likely to purchase next if you send them an email or offer them a promotion. Because these machine learning algorithms detect patterns that human beings often miss, it is able to serve a lot more intelligence to marketers a lot more efficiently and catch things that we wouldn’t normally find, at least not as efficiently as AI.


The ability of artificial intelligence to connect dots that would be tricky for humans to do, combined with it’s capacity to crunch numbers on a massive scale is an extremely powerful tool to possess.


eCommerce and Artificial Intelligence Cartoon with lots of products in a basket


AI Merging with eCommerce


We’ve all heard the rumours about being tracked around brick and mortar shops by cameras tracking what is in our shopping baskets, what are shopping trends are and even the route customers take around the shop.


Granted this is designed to make retailers see more but, whilst it is a bit (well, a bit more than a bit) of an invasion of privacy, we, the shopper, gain from it as well.


Customer experiences are becoming tailored to the individual which is great! No?


Personally, and I know I speak for many others out there, I want to spend as little time as possible traipsing around shops trying to find that illusive product, whether it’s cotton buds or ink pots, and if someone makes this easier for me then I will welcome it with open arms.


For some time in America billboards have concealed cameras behind the motorway billboards to track which cars have been exposed to the ad it’s displaying.


So when Brett Bushkin (for some reason this is the name that has sprung into my head when trying to come up with an American sounding name) drives past the ad displaying a package deal on some Jesus sandals and frog socks, when Brett later drives up to the store in his olive green Chrysler to follow up on the ad, the store knows that Brett is likely to have seen the its billboard and they can now build a detailed profile on Brett which it can then use to apply to others. The stores can machine learning system can now sort this data and roll out ads at the correct time of the day to when Brett, and those like him, are likely to see it.


This style of marketing has now been stepped up and advertisers can target specific cars with specific ads. Indeed a year ago Britain’s most famous billboard in Piccadilly square started doing exactly this.


The leading digital researcher Gartner predicts that by 2020 around 80% of customer interactions will be handled by artificial intelligence and in as many as 95% of all customer journey’s, artificial intelligence will have been employed at some point.


Should we be weary of this or should we embrace it?


AI disruption in eCommerce


Disruption is a loaded word and it is often used as the link between AI and eCommerce. To me ‘disruption’ undeservedly carries various negative connotations, especially when associated with eCommerce. Invasion of privacy jumps to mind.


So when it comes to AI and eCommerce, I don’t think ‘disruption’ fits the bill.


Revolutionising, enhancing and refining are words that should be sandwiched between Artificial Intelligence and eCommerce. We must accept that we live in a data driven world and when handled correctly it has the power to turn the mundane into the sublime.


Shopping experiences are truly bespoke and whether you want to spend an hour or a minute in shops, you will increasingly leave more satisfied as AI enhances these often tedious experiences.


On a personal level, as I’ve already mentioned, I cannot stand the thought of wandering around shops, either online or in real stores, looking for the colourful cotton boxer shorts that I’m particularly keen on, or what ever the product is… If I can be signposted to them more efficiently, or better yet them brought to me, then I’m going to be happy.


Over the past decade we have seen how much AI has changed the way we spend our time, most of the time unknowingly. However most people don’t appreciate that AI has been around for considerably longer.


Learn more about our leading digital Agency Here


Past Predictions – Where are they now?


Making predictions around technology has become increasingly tricky and, as Moore’s law demonstrates, the speed of technological advancement makes making predictions nigh on impossible – especially when thinking about timings. However, as the law also suggests, the ceiling preventing us from progressing further is also fast approaching and current human knowledge will not be able to enhance the machines we have created.


Over five years ago, when AI and eCommerce was a lesser-known beast, leading qualitative and quantitive analyses firm McKinsey Quarterly wrote an article that looked ahead at where this space was moving. The following is a version of one of their predictions


Near Field Communication


The rise in NFC (Near Field Communications)



Intrigued by his friends headphones, Brett taps his NFC enabled phone against them, they too have these NFC powers.



Brett’s phone then asks her to take a selfie.



In the app Brett can now see what he looks like with a pair of headphones and can select the colour.



Brett can now share the image with his friends who can vote directly on which colour they prefer/suits him best.



Brett now receives a message from Shopify offering him a free months membership subscription or a free premium plan if she buys the headphones



Brett loves the beige headset and decides to make the purchase. Yay Brett



When Brett opens the recently delivered box the next day he is asked if he wants to take a picture to share online online with his friends. The more shares the more perks he gets such as a Spotify subscription from, for example, his mobile phone provider who has a partnership agreement with the headphones.


Whilst all of the above isn’t completely familiar, generally speaking they got it pretty spot on and most of what is described is seen during our every day lives. Now where is all this going?


The Future of Ecommerce 


Where is AI and eCommerce heading?


In some cases we’re already there, in others we’re almost there and others are speculative, however the following is a list of a few areas where Artificial Intelligence heading:


  1. Albert – A trustworthy sounding AI led marketing platform.


Albert is a software that automates the campaign management of your paid and non-paid media campaigns. Making use of predictive analytics, machine learning, language processing and feedback, Albert keeps a check on campaign execution, optimisation, analysis, testing and insights.


Now for those businesses who have neither the time or the inclination to run there own marketing campaigns themselves Albert might be a good alternative to the expensive agency fees that they would otherwise be facing.


  1. Boost Conversion – stop those slippery customers from getting away


As more and more data is being collected on customers we’ve seen a rise in the number of AI driven strategies aimed at engaging these customers and helping them to convert



This is one of the big players in conversion marketing today and the statistics out there are fairly irrefutable. For example Kia introduced their digital friend Kian (a Facebook messenger chatbot) to their website in late 2017. Kian is programmed to answer any question a user might ask and if Kian didn’t know the answer guess what, through machine learning, the next time he will. Since the Chatbots introduction Kia reported that it had tripled its conversion rates in the space of 4 months – and something tells me this wasn’t down to a ground breaking new car design. From improving customer service to shortening the sales steps, to up-selling and cross-selling, chatbots have revolutionised digital marketing in recent years.



Retargeting is a key part of digital marketing as it keeps your brand in front of previous visitors to your site thus increasing the likelihood of them converting. Shoelace is an example of an AI tool designed for retargeting. With some leading agencies estimating that only 2% of shoppers convert the first time they are on a store, the benefits of keeping your brand at the forefront of your customers as they move around the internet is invaluable and can save a considerable amount by avoiding targeting the wrong type of customer.



Time is of the essence in this increasingly fast moving world we find ourselves in, and if you’re able to address customer reviews as soon as they make them then the chance of conversion can be dramatically increased. Through machine learning and AI, Keatext is able to sort through positive and negative reviews and pass them onto you in real so that you are able to address any issues head on and even create a ‘feedback loop’ so that you are able to display a more relevant product to the customer straight away.


  1. Content Recommendation


As we know creating engaging content is a key factor to driving traffic to a site and boosting your page/site authority in the eyes of search engines. Many companies are now experimenting with bots that not only suggest potential trends but also analyse and provide feedback on the likelihood of that content generating shares or engagements. For example Rocco AI is a social media bot that recommends more attractive content. Rocco can even be integrated with Slack where it can analyse a brands tone of voice and make recommendations for original content ideas.


  1. Image Recognition – shining a light into the shadows


For years bots were unable to see or understand what an image was however, times a changin. Google, Facebook, Amazon and every other big digital player have developed bots that recognise what an image is and can provide a detailed description of what the subject matter is. So for instance, if you take a photo with you smart phone you can then carry out a search on that image and you will be delivered products that are similar to the image in the results.


  1. Voice Commerce – “Talking to oneself – the first sign of madness”


Now if speaking to robots isn’t your thing then you better change your tune because there is absolutely no denying the direction we are moving in.


Voice commerce is the interaction with an eCommerce site when there is no screen to engage with – Amazon Echo, Google Home etc. This is great for repetitive orders or things that you have already put onto your shopping list and like it or loath it, we are getting more and more comfortable talking to machines and inevitability this trend is going to continue.


It hasn’t quite reached the level of product search or product comparison, so there is still a need for traditional eCommerce experiences and this will not change. However the way voice search will augment how we shop by providing greater convenience is certainly where we are going to see a lot of growth over the coming few years.


  1. Combatting Fake Reviews


Reviews are an essential part to modern day shopping with Shopify estimates that as many as 63% of customers are more likely to make a purchase from a site with product ratings and reviews. More often than not you’ll be able to find an alternative product for a similar price on a different website and people’s decision as to which to buy is largely based upon a star rating or product review as well as expense of course. However, who is writing these reviews? Well, it could be real customers but increasingly over the past decade machine learning AI has evolved to be able to write reviews that are nigh on impossible to tell are not written by a human. This being the case how can we combat it? Well, through AI obviously.


Whilst it might be tricky for us to decipher between fake and real, AI is being developed to spot behavioural patterns in reviews that can indicate whether or not the review is genuine or not and remove or flag up those that is suspects to be fabricated.


The Dangers of AI


But Alas – That Which Glitters is Not Always Gold


What are the negative effects of AI? I admit I have been waxing lyrical about artificial intelligence and how it is going to help me minimise the time I spend in shops, which is great. However of course there are consequences; some more concerning than others.


In the last few years there are several embarrassing examples of people being caught out. Take for example Microsoft’s chatbot Tay being provoked into becoming a racist back in 2016 when social media users bombarded it with confusing quesitons. Within a day Tay was tweeting about anti-semitism, racism and dictators, to name but a few.


But this isn’t eCommerce I hear you say. Well, true. However the principles of manipulation still apply and depending on what questions are asked to your eCommerce chatbot, its effectiveness and the results it returns can be varied. This can have a devastating impact on your brand.


Useful AI is based almost entirely on the quality of the data that is collected meaning consequences could be disastrous should their be discrepancies in the data from which AI is learning. JJ Guy, the CTO at Jask – a leading AI security provider, said: “The algorithms themselves are easy. Collecting, classifying and labelling datasets used to train the algorithms is the grunt work that’s difficult – especially datasets comprehensive enough to reflect the real world.”


  1. You might end up living a sheltered existence


Whilst it is great to be presented with your colourful boxers that you had been searching for as soon as you land on a page, think of all the things that you might be missing. Whether it is ‘Recommended For You’ Netflix films or ‘You Might Also Like’ popups that appear across countless eCommerce stores including, and perhaps most dangerous of all, publishing and news sites, gradually someone else – or something else – is tailoring our lives for us.


Our ability to choose is being diminished and this is dangerous on so many levels, not least that our lives will become blinkered and driven by what brands think will make them the most money.


  1. Anything you can do I can do better


Artificial Intelligence doesn’t bode well for the job market as, as we have pointed out in this article, machines are increasingly becoming more efficient than humans, which ultimately means that people are going to start losing their jobs.


The Future is Bright, the Future is Artificial Intelligence


It might sound obvious but Artificial Intelligence is here to stay. From self driving cars to fully customised eCommerce shopping experiences, soon enough every part of our life will involve sound element of artificial intelligence, whether we are aware of it or not.


In 2019 expect to see an upshift in eCommerce stores implementing AI technology to help drive sales. In fact I would predict that by the end of 2019 the vast majority of online merchants will be making use of AI somewhere on their site, whether it is voice powered shopping, virtual fitting rooms or computer-generated assistants offering a hand.


My only advice would be to try and avoid being hit by a self driving car and remember to take shopping online with a pinch of salt and recognise that there are some very clever machines out there getting us to open our wallets.


As a leading digital agency Diffusion Digital would be happy to answer any questions you may have – feel free to get in touch through the form below or alternatively you can call us on +44 (0) 203 141 2000