In the third episode, we saw how Artificial Intelligence can interpret a fashion customer’s purchasing desires in real time, even when they don’t yet know what they’re looking for. But detecting intent isn’t enough: you also have to present the customer with products that will spark their desire, and then reassure them enough to take the plunge.
In this fourth episode, we explore precisely these two challenges. First, how companies like Zalando use hyper-personalization to suggest the right products at the right time and drive purchases. Next, how ASOS addresses one of the main barriers to online shopping, not being able to try on the product, through its Fit Assistant.
The customer doesn’t know exactly what they want to buy during a shopping trip. To make them want to buy, they need to see the product to be able to say «hey, that’s not bad!»
As with physical commerce, you need to use clothes displays and models to present the products that are most likely to please them.
Having observed a customer’s behaviour in real time and detected their individual tastes, you can reap the benefits of hyper-personalisation: being able to suggest products that match the tastes of each one. This is particularly true within the search engine, where the customer does not have a firm idea of what they want, nor how to express what they want from generic searches.
Hyper personalisation will enable the top line display of the products most likely to be of interest to each customer, according to their buying intention, even from single word searches. Thanks to instantaneous and hyper-personalised word suggestions in real-time, the search engine can help the customer find their words, refine their search and find products that meet their needs.
What’s more, product suggestions enable the entire buying journey to be hyper-personalised, thanks to cross selling logic. When viewing a product sheet, the customer will be offered products related to the items viewed, matching their tastes and buying interests.
Having established that customers engage more easily with entire outfits than single items, the biggest players in the industry go even further by suggesting complete styles, based on the behaviour of the customer. Entire outfits are proposed to each customer, adapted according to their searches, the products they have viewed and the analysis of many other variables.
Implemented at Zalando, this practice allows the customer to discover new products that are right for them. It also has an impact on the site’s performance: 40% increase in basket size and two times more conversions than with single products.[1]
In the following example, we look at two shopping journeys that took place at Zalando. .
- In the first sales journey, I’m most interested in finding trainers, in «streetwear» style .
- In the second, I’m more interested in finding work shoes, my usual style being business casual.
We can see that, from the first interaction, my purchasing intention is analysed and used in real time. The entire purchasing journey is then hyper-personalised: product suggestions, search results, product rankings are adapted based on my behaviour and my preferences which are detected and refined through sessions.
Even though we can now advise each customer and influence the route for every purchase, there is still a barrier to completing an online purchase: the inability to touch or try on the product.
Online it’s often difficult to know if the size chosen by the customer suits their shape, and if the product will have the quality they expect.
Again, artificial intelligence en - ables us to address this pro - blem. Let’s take the example of Asos. The Asos “Fit Assistant” allows the customer to pro - vide a series of details about their shape and size: height, weight, age, chest – in order to match the customer with the right size in each product.
This is a feature that has had dual impact: the conversion rate with Asos Fit Assistant has in - creased by 11.5% and returns have reduced by 4.4%.[2]
In our example, I decide to buy t-shirts but am not sure which size to choose, so I enter in - formation into the fit assistant information: I am a six-foot 26- year old man, weighing 80kg, of average build. I also provide in - formation about the sizes I wear from my favourite brands and my clothing style; whether I pre - fer slim or relaxed fit.
The algorithm analyses my shape and size, preferences and the product sizes. It ad - vises me to take one of the t-shirts that I’ve chosen in large and the other in medium.
Zalando to spark desire, ASOS to resolve sizing concerns: two complementary approaches that demonstrate just how hyper-personalization can transform every stage of the impulse-buying journey into a conversion driver.
In the next episode, we’ll shift gears completely to explore the world of deliberate purchasing, using the example of home furnishings and DIY. This is a purchase that requires planning, involves very real constraints, and demands a very different kind of support. Join us in Episode 5 to find out more.
[1] Source : Fashion United
[2] Source : Fit Analytics
Crédit photo : ©Shutte