How can AI-powered personalization drive revenue for retailers?

Did you know that Amazon’s product recommendations on site drive 35% of their sales? The revenue-generating potential of real, individualized recommendations is massive. 


Over the last few months we’ve all read about the speed at which digital is progressing. Apparently, more progress has been made in the last 5 months, than in the last 10 years. Almost every retailer is talking about digital transformation. Digital is now firmly cemented in the minds of business leaders as a major priority. From here on out digital is imperative.To me, this goes hand in hand with personalization.


Personalization is no longer an option to explore or a nice to have, it’s a strategy that needs to be implemented if you’re going to keep up with, or outdo, the competition. We’ve been catapulted into a time where physical face time with customers has decreased drastically. Personalized digital touchpoints will become the most important brand-customer connection. 


Ecommerce is the ‘new normal’ (I hate that term but jumping on the bandwagon), email has always been an incredibly important channel, but I’d go as far as to say it is now even more important. Email is the shop front to your brand in the inbox, and therefore, your personalization efforts need to be accelerated in the inbox, as well as onsite, to ensure customers are kept interested, engaged and buying. Brands need to deliver value through email and site personalization. I’m not talking about segmentation – say 1:10,000 comms, I’m talking about truly individualized content experiences for customers.


Many brands are comfortable at the ‘micro-segmentation’ level of email ‘personalization’. They can use dynamic content or content automation tools, or maybe even use collaborative filtering to show customers content based on a preference – perhaps a last browsed category or a brand preference. However very few brands are at that individual experience level which is surprising when we know that personalization has such a positive impact on the entire business.


The personalization basics are easy, and most brands have got them in hand – first names, birthdays, loyalty points, order info etc., but brands need to do much more. They need to open their inventory to start to personalize products in a real way. Whatever stage of the journey a customer is at (discovery, view, purchase) you can elevate their experience at every touchpoint to help them chose the product they want, they might not even know it yet. This is where you can really help them. Having too many options can cause friction and drop off. Here’s how you can assist customers when they’re making purchases.


Catalogue size is increasing as businesses expand – take fashion retailers for example. Many of these now sell beauty products and have homeware ranges alongside their core fashion product offering. When there’s more products in the catalogue there’s less visibility for each product. On a site with 100,000 products a customer might buy 5 products in one session. Perhaps on a daily basis 90% of products get viewed; as the catalogue gets larger only 70% or 60% of these products might get viewed. Consider this too: 70% of users don’t scroll to page 2 of products ( This is where a personalized strategy can really take it up a notch for you (if you want more in depth information on this topic, check out this article.)


Open your inventory and use AI technology to help your customer access what they want to find. Surface 'similar' items and recently viewed products on webpages and in email to make their life easier. If the customer is looking for a black t-shirt show them 5 other ‘similar’ black t-shirts. If the customer wants to jump back to the first black t-shirt they viewed, there it is in the recently viewed items section of the page. Help customers to find what they want easily using AI.


If customers aren’t browsing through all the pages on your site, use AI to boost the products that they’re likely to like. Include AI-powered product recommendations in email and on site. Kickdynamic can help here, we’ve built a revolutionary AI prediction engine (LYNX) that can be used to power product recommentations in email and everywhere else. Use AI to understand what the individual likes and display recommendations where they're looking. Leveraging the same AI-powered personalization on-site and for email guarantees synergy between the channels. 


Of course, there’s some getting to know you time required; that first visit and the few products the customer views at the first touchpoint is important – it’s where the personalization starts. Each time there’s interaction you learn more about the individual and are able to personalize more and, make better recommendations as they come back. The goal is to enable customers to discover products – the right products – each time the engage with your brand so that they buy. The goal is to re-acquire the visitor at every touchpoint. Personalization is a very sticky way to keep customers invested in your brand; keep showing them what they’ll like, and they’ll keep coming back. They'll see more they like, so they'll likely buy more. They'll be more engaged with your brand so they'll buy more frequently. Stats suggest they'll spend more too so expect AOV increases too. 


Remember that you can have an element of control – that’s really important for conversion. With tools like LYNX you can chose to boost / filter products to influence conversion. Decide to only recommend products that are above a set price for example.


Once you understand the customer’s interests (brands / categories / price points / products viewed / added to cart / purchased) you can use this information universally. Another way to successfully retain a customer is to cross-sell. Use global site data and AI to recognise if somebody has bought a shirt and then recommend jeans, a jumper and some shoes to go with it. That's a strong behaviour influencer. 


Cross-sell is incredibly powerful and needs to be better leveraged, particularly in the email channel. Integrating all of the ideas posed in this blog in email must be a priority. At this point, email is the virtual store front for online retailers. Email content should be individualised to appeal to each individual. It’s much more likely that each customer will like the products in the email if they’re just for them, rather than picking 6 products that mostly likely will only be liked by a few recipients. 


Product personalization on site and in-email is actually about bringing the value to the customer, not to the brand (that's a nice by-product). Value to the customer is the first priority. Don’t keep selling customers the same thing. So many brands get caught in a cycle of promoting the same small group of products to everybody. Maybe this is down to tastes, designers, commercial teams, the big bosses. No more. Content should be determined by the individual doing the buying, not the selling! To truly add value to your business you need to consider the unique profile of each individual customer. To do this at scale you need an AI-powered prediction engine.


Kickdynamic book a demo

Back to our blog

Let's connect:

Follow us on Linked In   Follow us on Instagram   Follow us on Twitter

Recent Posts