Summits Yellow

Spotlight: Sentient Aware

David Murphy

IMG_2174The majority of search is text-based. You enter the search term, hit enter and hopefully get what you want. Even with the move to voice, the concept is the same; it’s just a different way of inputting the search term. This works well enough for most people, and one company in particular has made a few bucks by focusing (initially at least) on this as its core offering.

The only problem with this is that sometimes you don’t really know what you’re looking for. Or maybe you think you do, until you see something better. And so various visual search platforms have emerged to offer a more, er, visual alternative. Cortexica’s Find Similar platform, used by Net a Porter and John Lewis among others, is one such example. It enables a shopper to take a photo of an item they like and use this as the basis of a search for similar items from the retailer’s inventory, based on elements such as colour, pattern and shape.

Another interesting take on the idea is Sentient Aware, which was demoed to me a few days ago. With Sentient Aware, the idea is that the shopper narrows down what they are looking for to a certain extent, then leaves the retailer to suggest items that might fit the bill, including some they might not have discovered using a traditional keyword search process.

Here’s how it works. Let’s take an online shoe retailer. On arriving at their website, the shopper is initially required to select the type of shoes they are interested in from a list. Once they have made this selection – ‘Heels’ for example - the Aware tech kicks in.Shoe stitched

On the next page, the shopper sees a banner bearing the text ‘“I’ll know it when I see it” Show Me Shoes”’. Or whatever the product in question is. When the user clicks on the banner they are invited to enter some basic information. In the case of shoes, it’s size, then width then type. Here’s where it gets interesting. Because the ‘Type’ selection brings up 10 different types of women’s heeled shoes, together with a question: “Which of these is the closest to what you’re looking for?” This is the retailer’s top-level view of their women’s heel range. The retailer can specify the number of shoes to display, and can change this number when the site is being accessed from a mobile.

So the user clicks on one of the shoes. On the next page, the shoe they click on is identified and the shopper is invited to click on a button to view the product details, or to click on one of 10 similar (but different) shoes to see more shoes more like the one they are clicking on. If they click on one of the 10 different shoes, the process is repeated again, and again, so they are able to get a visual on a lot more shoes than they would be likely to see otherwise. Which obviously works in the retailer’s favour as well.

In fact, Sentient says that Shoes.com, which has deployed Aware on its Canadian site, went from exposing 20 per cent of its catalogue in any given month prior to deploying Aware, to 93 per cent in the first month after deployment. The same customer also saw a 20 per cent in conversions on desktop and a 33 per cent increase on mobile. Basket size increased by 4.5 per cent on desktop, and by 7.5 per cent on mobile.

In addition to Shoes.com, shoe retailer Skechers has also deployed Aware on its sites in the UK, US and Germany.  If you want to try Aware for yourself, click here.