Skip to main content

MagentoLive 2018: how to do AI in Magento

Feedback from MagentoLive Europe – 3rd part

IA, Big Data and Magento

Where is this leading us? To the inevitable threat of the year, AI. Hosted by Isaac Moshe, the Chief Marketing Officer, the presentation was organically structured and well paced enough so any bystanders’ stepping in wouldn’t be much of a hassle. As a regular developer this talk was an eye opener as to how AI can be integrated and used practically within work in a Magento specialised web agency.

For starters Isaac made sure everyone was on the same page. He narrowed down what AI is in retail, moving aside general AI to promote specific AI. In his words, “artificial intelligence is a solution that uses data to adjust outputs towards a given set of goals”.

But what kind of data can be used? For the past decade and a tad more web developers have been used to work the data, but most of us did it with the starter kit: transactional data. Though this type still is relevant, as an important part of the customer experience’s DNA online, it clearly belongs to the past.

Indeed, 98.4% of the data that can be collected and used is the behavioural data and this is the one that matters the most today and for the future. Isaac gave very straightforward numbers for those who struggled to get the picture : 80% of the online retailers believe they are providing a customised enough experience, while only 8% of the actual customers, the people, agree. The gap pretty much shows the extent of work tech companies have to deliver to online shops, separating the noise from the signal in order to yield superior results.

If the need to make the move doesn’t feel strong enough yet, consider a worldwide known example of how to use AI the right way. Netflix, in the rearranged outdated words of its Director of Global Communication Joris Evers, offers 130 millions of different versions. Isaac presented the crowd how analytics are used by Netflix, dissecting user data and using ML algorithms to make the most of it.

We’ll leave that part with a last number and the obvious question it raises : today’s predictions assert that by 2035, 59% of online retailers will use AI integration.

La personnalisation grâce à l’IA avec Nosto

That first part of the talk was very much focused on big data but of course, Nosto being an actor in the field and a sponsor for the event, there was going to be some bridge tossed across to reach developers. Isaac took the demonstration to the Magento field, which sells more physical articles than virtual ones.

He introduced us to the concrete application of AI in a customer’s experience, showing how different people would get different content, less based on what they requested and very much more on what AI determines is of interest to them. This aspect was echoed in the keynotes and other talks but again, the demonstration was through concrete cases : AI enables us to provide tailored, curated advice, to guide the customer to a better experience.

The talk was concluded by Neil Bruce, from Toolstop, and Rory Gilmore, from Nosto, who explained what challenges Toolstop was facing and how Nosto’s AI integration helped solve them. From the study cases it was clear that AI was not just a way to customise experience, but also the relationship between a customer and a retailer.

Thank to it, you may forget wide unprecise coupon campaigns and offer the incentive to those who really need it. You may adjust other incentives and also propose more relevant advertising and, why not, customised VIP content.

Sophie Bertrand - Directrice commerciale Kaliop Digital Commerce