Visual Search Needs To Be Fashion Aware To Understand Consumers

They say a picture is worth a thousand words and research suggests visual memories are the strongest form of memory. It is always easier to explain something if we have imagery rather than just using keywords. This is one of the reasons why fashion retailers are enthusiastically introducing Visual Search to their consumers.

Visual Search is a feature that allows a user to upload an image, and find matching products from the inventory.

Visual search is revolutionizing how consumers are discovering and purchasing merchandise. Currently, ASOS allows shoppers to purchase using image search, and Neiman Marcus reported a strong quarter after developing the technology as well. Companies like eBay, Ted Baker, Blippar, EasyJet, Levi’s, and Disney are also experimenting with visual search.

“The future of search will be about pictures rather than keywords,” stated Pinterest CEO Ben Silbermann. The Global Visual Search Market is estimated to surpass $14,727m by 2023 growing at CAGR +9% during the forecast period 2018–2023*.


Visual Search is geared towards the tech-savvy generation who hold thousands of images on their phones. Each image is someone’s fashion statement. This technology allows users to be inspired by any image in their phone, search for that item and buy it with just a click.

But… is Visual Search user-conscious?

Understanding your user plays a huge role in making Visual Search successful in the fashion industry. The technology in itself is simply not enough to guarantee consumer and retailer success.

The challenge (especially for fashion retailers) is that the system fails to understand the user. Typically, the same criteria used in searching for non-fashion items, such as household furnishings or kitchen appliances, are used to display results, rather than fashion specific results.

So, what does a customer want, and what results are they expecting?

Asking these questions is crucial for retailers, as more consumers are prioritising experiential retail journeys than ever before. Streamlining these results will only enhance that shopping experience.

A recent survey showed that 78% of respondents expect to find ‘something similar’ and not the exact product when they use Visual Search. So what this tells us about the customer requirement is that Visual Search technology does not have to find the exact same product every time, but it must understand the important characteristics of the product. The technology then needs to predict what the shopper might be looking at from a fashion perspective to deliver a better user experience.

Digging deeper we found that the majority of respondents said that as their first priority they would like to see visually similar products within the same apparel category. For example, if a women’s top is uploaded, they would want to see similar tops/t-shirts/shirts, regardless of neck style, sleeve length or any other attribute first. Secondly, they would like to explore products with similar attributes. For example, if a polka dot women’s top is uploaded, then the user expects to find tops with the same polka dot print. Lastly, the user expects to find the exact same product as in the uploaded image.

What does this mean for retailers?

When a single brand fashion retailer, incorporates Visual Search, it is not a question of finding the exact match in a comparatively small inventory, it is about finding something similar, in terms of colour, pattern, sleeve length, or other attributes. Using this tiered prioritization, the retailer can expect an increase in product views, clicks and PDP (product display page) contribution, leading to increased product discovery and user engagement.

With a multi-brand retailer or marketplace, owing to their large inventory, their system can prioritize the exact product in the image, followed by visually similar results. This will lead to faster product discovery, a decrease in search time, and an increase in PDP to conversion ratio with faster checkouts.

-

Streamoid provides user conscious Visual Search solutions for fashion retailers. The article was co-authored by Varshitha and Murtaza from the Visual Search team.

Varshitha — Fashion Associate, and the voice of the consumer, championing collaboration between technology & fashion, manages visual search and vision-based engagement products

Murtaza — Product Lead, with senior responsibility across the majority of Streamoid’s product range.

References: