Pinterest search gets visual
Pinterest is an incredible visual resource, with over 100 billion ideas saved by 175 million people all over the world. Food is the most searched category on Pinterest, as well as being the content most frequently marked ‘tried’ – when users say they have completed a recipe or activity and leave comments about its success.
In February this year the company introduced Lens, a visual search and image-recognition feature which enables users to find content using an image rather than words. Snap or upload a picture to the platform and Pinterest will return similar designs and other related items; they claim the technology can currently recognise 1 billion objects.
But the search does not just return visually similar objects. A picture of a pair of trainers will find similar shoes, but also outfit ideas that match the trainers. You can even aim your phone at the night’s sky, for pins related to stars and other things extra-terrestrial.
Initially rolled out for fashion, home design and raw ingredients, this week Pinterest announced yet another form of visual search – show them a picture of what you’re eating and Pinterest will suggest matching recipes.
Advertisers will ultimately be able to match their ads to these visual searches and to interact with consumers showing an interest in their products, all the way through to the ultimate purchase, which can also be completed within the app.
Their food-related text search has also been overhauled with improved filters – refine your results by dietary preference, preparation time, even what you have in the fridge. Even with search turning visual, these developments emphasise the importance of keyword-rich metadata for images, as well as the need to enable Rich Pins for your site’s food content; these pins contain vital extra information (ingredients, preparation time, etc.) that will help users find your client’s content.
Google is also keeping up with developments in visual search with its own product, also called Lens. Also powered by smartphone snaps, this search function (still in development) is focused on the outside world (monuments, street signs, etc.), as opposed to Pinterest’s targeting of clothes, food and furniture.