The Value Of Visual In The Multichannel Customer Journey
Consumers preference for visuals is being reflected more and more through their online behaviour and will be ever more present. According to an eMarketer study, around three quarters of U.S. based internet users regularly search for visual content prior to making a purchase. Pinterest has also reported 93% of its users use the site to plan purchases, which led them to develop tools to create searchable images where users can tap just one part of an image to find product information.
Recent research from SmartInsights shows 86% of people believe that images are more important than text when it comes to buying clothes and 85% agree when purchasing furniture. Not only that, 90% of information transmitted to the brain is visual and our brains must work harder to process words. Because of this, on average people only read 28% of the words on a web page with 74% of consumers saying text based keyword searches are inefficient in helping them find the right products online.
Here, Michael Wood, Principal Digital Consultant at Astound Commerce UK, explains why visual search is so important for retail in 2020.
Mobile’s part in driving visual search
Mobile has a large part in helping drive this behaviour. Image based search and display is preferable to text on the smaller screen and customers have grown accustomed to the discovery experience offered by the likes of Instagram, using these as their source of inspiration rather than browsing through the large online catalogues offered by online retailers. Savvy businesses, especially those with a high social media customer base, have started to adopt similar approaches, utilising image search tools in mobile apps.
Fashion retailer ASOS launched StyleMatch last year, initially in selected markets. ASOS needed to solve what they perceived as an ‘upstream’ problem, with visitors using their mobile app only discovering a small percentage of the catalogue due to screen size limitation. Now, customers can upload a photo or screenshot and search across ASOS’ 85,0000 products for similar items. While ASOS have not released any numbers on how well StyleMatch has done, they have seen enough to release it across all their markets.
And just this month, Amazon has rolled out its AI-powered fashion search tool to UK users of its mobile app. StyleSnap is a tool inbuilt into the wider Amazon App that uses machine learning to allow customers to shop by uploading images of products they want and be matched with similar products available on the retailer’s marketplace. The feature currently works for searches on dresses, tops, bottoms, shoes and bags and supports both women’s and men’s clothing.
For more than just fashion
It’s not just fashion retailers looking to take advantage of visual search. Wayfair’s Search with Photo tool allows customers to upload images to the website and scans Wayfair’s 8 million SKUs to see if they have something similar. A recent update automatically selects crop areas for search and identifies multiple items in each picture. When the update was release, there was a 58% lift in repeat engagement with the tool.
Algorithms and data analysis
Another area helping drive image search adoption is the constant improvement of the algorithms through data analysis. Alongside Pinterest and Instagram, Snapchat, Google, Microsoft and Facebook are all refining their visual search offerings, learning from the mistakes of the past (Google’s well publicised issue with detecting humans vs. wildlife being one) using data taken from user behaviour and customer feedback. Even so, there are still improvements to be made.
Visual search is challenging for even the big tech companies. If it was as easy as looking up and returning a few key points e.g. shoe, men’s, black, the results would be no better than the basic search results you get from most ecommerce engines.
It’s the additional elements such as style or function that add complexity but are necessary to ensure what the user sees is concise and relevant. There are also the photos themselves, with user generated content varying in quality (blurred images taken at odd angles etc.).
Optimise images for visual search tools to easily process
Even if a retailer is not ready to add visual search to their app or website yet, there are still other considerations for online search. SmartInsights recommend that sites should still ensure images are optimised using structural data and other traditional SEO tactics but that going forward, imagery should be clutter free and easy for visual search tools to process. Other suggestions include:
- Offering a range of clear images for each product
- Optimising image titles with target keywords
- Submitting image sitemaps
- Setting up image badges
- Optimising image sizes and file types
- Running structured data test
With Gartner suggesting that by 2021, early adopter brands that redesign their websites to support visual and voice search will see an increase in revenue of 30%, it’s certainly worth considering where visual search sits in your road-map.