Fashion has been the muse for many technological breakthroughs. It is the reason why Google Image Search was born in 2000, when the most popular search query was J.Lo’s slashed-to-the-navel green dress that she wore for the Grammy Awards.
Fast forward to today, fashion is currently inspiring machine intelligence companies to develop visual technology that can take visual commerce further and increase conversion rates for online fashion retailers.
The rise of visual commerce
A previous Omniconvert article was describing the rising trend of innovative visualisation methods and personalised looks used in fashion eCommerce one year ago. And they’ve only become more predominant since.
Professional photography in fashion eCommerce is setting a higher and higher bar for visual content on online shops.
The most notable emphasis on photography was made recently by Amazon when opening a huge photo studio in London to feed its ambition of becoming the best place to buy fashion online. And let’s not forget fashion brands like the $86M-funded Stance that differentiate their eCommerce platform through outstanding lifestyle photography.
User-generated photography is also here to stay, as buyers appreciate visual inspiration from fellow buyers, considering their photos of products more authentic and trustworthy.
Clozette, a fashion social network, is curating Instagram photos of fashion bloggers and buyers and posts them on their platform for inspiration. And they are not alone – Asos, Burberry, Coach, Desigual and many more have similar implementations.
As exploratory search is more predominant than navigational search in fashion e-commerce, visual content is only here to stay, and fashion will continue being at the forefront of visual commerce.
But there is something more to visual content than just pleasing the eye.
3 ways fashion e-commerce can take visual commerce to the next level
As machine intelligence companies managed to unlock the content in visuals with visual search and image recognition technologies, the fashion eCommerce industry can now benefit from shoppable visual content, mobile image search, and smart recommendations and personalisation – all leading to higher conversion rates.
1. Shoppable visual content
Every photo out there that your buyers get inspiration from can easily lead them to matching or visually similar products on your fashion eCommerce platform. And this can be done without any manual work and pre-setting, just by using visual technology to automate and scale.
The previously mentioned fashion social network, Clozette, has taken their curated Instagram OOTD photos to the next level, placing a Buy Similar button on every image. Buyers can simply crop the image to focus on the product they want and see what similar ones Clozette has to offer in their shop.
This doesn’t have to depend on existing curated photos, it can as well be implemented as an image upload search. Basically, buyers can upload any photo they want and see if they can find matching or visually similar products inside the online shop.
2. Mobile image search
Smartphone cameras can now bridge the gap between the offline and online worlds when it comes to shopping for fashion.
As visual technology made it possible to identify products in photos and search for similar items, more and more fashion retailers started enabling mobile image search on their m-commerce platforms.
When eCommerce conversion rates on mobile are still lagging compared to on laptops (0.96% vs 2.71% according to a Monetate research), smartphones can now offer a unique advantage through this offline to online bridging by visual tech.
Flipkart, the largest online retailer in India currently valued at USD 15B and one of our customers at ViSenze, is one of the mobile image search trailblazers. They allow their mobile users to point their cameras at any fashion item they like around them and find similar ones in their shop.
In certain situations, people would want to find a nature-inspired printed outfit or a shoes to match the exact colour of a bag they have – there are no limits.
Sometimes buyers don’t want to browse and explore, they just want to find something like what they have under their nose.
We made an experiment to compare keywords search vs visual search and we noticed that people give up on searching for a specific item in an online shop after 1 min and 30 seconds on average, with a fail rate of 96.6%. Visual search on the other hand can allow people to search by images and spend only 10 seconds from taking the photo to finding a matching or similar item.
The quality of the results obviously will depend not only on the strength of the technology, but also on having similar items in stock.
3. Smart product recommendations and personalisation
Sometimes buyers might like a product but would prefer something just slightly modified. Sometimes they really like a product that is out of stock. Product recommendations based on visual similarity in terms of colour, style, pattern and shape can come to the rescue for both the buyer’s satisfaction and the eCommerce revenues.
There are different ways to implement visual technology for product recommendations: You May Also Like (displayed for every product that is viewed) or Find Similar (displayed on demand when browsing).
The position in the page also impacts the conversion rates, with the best results – an uplift of 50% conversion rate – being witnessed by one of our customers that places similar items on the right side of the product image and details.
The same technology can power smart personalization, by highlighting on subsequent visits products that are visually similar to the ones previously viewed or purchased. This can also pair nicely with other data points like weather conditions or geolocation that Omniconvert is providing for segmentation.