GLOBAL: Visual search is rapidly becoming a mainstream proposition among younger consumers, and while associated ad products are not keeping pace the longer term holds out the promise of “visual commerce”, according to a new WARC report.

The WARC Trend Snapshot: Visual search, outlines how this particular capability has benefited from recent advances in artificial intelligence and machine learning, with the trail being blazed by Pinterest.

The image-based social platform launched its Pinterest Lens product in February 2017 and 12 months later reported that users were conducting 600 million visual searches every month.

But even as the development of visual search systems continues apace, use of the technology for advertising purposes remains stubbornly rooted in the formats developed for text search.

Most visual search ads currently purchased on Pinterest, for instance, are carried out with parameters familiar to those who buy Google AdWords, based on whether the query leads to a broad or exact match.

“People love being able to go into a spreadsheet to figure out their AdWords campaigns,” explained Clark Boyd, a digital marketing consultant and visual search expert.

“You can’t approach it in this way [in visual search]. It is hard to commoditise in a way that both computers and people can understand.”

If a consumer takes a photograph of a sneaker, for example, the visual search algorithm must make judgements on tangible factors such as style and colour, as well as the intangible, including trends and individual intent.

“That is where I think it is tricky to monetise, because the commoditisation of these ideas isn’t as simple as it is for text,” said Boyd.

But as more and more younger consumers turn to visual search – 51% of 18-24 year-olds in the UK have used it – observers expect the path to purchase to be reinvented yet again as everything becomes buyable.

Oliver Tan, founder of Singapore-based technology provider ViSenze, envisages the rise of “visual commerce”, with users able to click on images on messaging apps such as Facebook Messenger and WeChat, prompting chatbots to supply relevant results and purchase options.

Brands then have to consider whether they want to collaborate with the likes of Google and Amazon, where their products may appear alongside cheap imitations, so decreasing brand value.

“This is one of the reasons that brands prefer [incorporating visual search into their own websites and apps] and social media provides powerful channels for them to do it,” said Tan.

Sourced from WARC