Since its inception, Pinterest has largely been used as a discovery tool for everything from home décor to high fashion. It features multiple ways of helping users discover new content, such as by recommending similar boards after an item is pinned, or through a search tool that helps you continually explore and refine what you’re looking for with related keyword suggestions.
And yet, it’s a highly-visual platform centered on images. So, in order to help users “find all those things you don’t have the words to describe,” Pinterest has announced a brand new functionality: image search. The concept behind visual search is to help pinners identify things they love on Pinterest, even if they are hidden within a Pin. Not knowing the name of the item is no longer a problem.
While visual search has been available via Google and Bing using similar AI technology, its application within Pinterest and other e-commerce-focused sites like Shoes.com opens up some very interesting opportunities to help users find and refine what they’re looking for with more direct commercial intent.
How Pinterest Visual Search Works
Pinterest’s engineering team is combining visual cues as well as the descriptions in Pins, or the “richly annotated dataset of billions of Pins curated by Pinners,” to produce visually-similar results.
Say you’re scrolling through Pins and spot a lamp you love in a living room shot – often the products within a pin aren’t specified, and searching for “lamps” within Pinterest will typically bring up a broad list of options. With visual search, pinners can tap the cropping tool in the corner of the image and highlight the lamp within the image. Pinterest will then provide similar image results as well as text refinement options, in many cases giving the user more details on the product as well as “where to buy” via Buyable Pins.
Pinterest also allows you to upload your own photos to Pinterest to initiate an image search. For instance, you can snap a quick picture of a cute pair of shoes, the perfect couch, or a fun cocktail. After using image search, you’ll be serve with similar pins as well as relevant categories to continue your search.
This is currently only available via the mobile app.
What This Means for Advertisers
While visual search will only serve organic results for the time being, we expect this to boost exposure on Buyable Pins since the visual refinement will display results only relevant to the specific item. Paid opportunities are expected to arise in the near future.
Pinterest’s visually-focused interface makes its image search tool more commercially-relevant than similar capabilities on existing search engines. Pinterest is already a top-of-mind platform in fields like home décor, fashion and planning for major life events (i.e. wedding, new home, etc.). For these relevant verticals, we see the potential for Pinterest to keep a larger share of its user base within the site when looking for more information visually rather than going to Google. It can mean Pinterest gains a bigger foothold as a search and content discovery destination.
Pinterest’s “deep learning” text refinement results (the categories above image results) could drive an evolution in keyword research – for both paid and organic. For example, when Pinners do an image search of a particular dining chair, Pinterest categorizes that particular chair with “quilted chair,” “tufted dining chair,” and “Parsons chair.” This leads to more granularity for deeper searching both in and out of the Pinterest interface.
- Pinterest is likely to not only better keep users within the site or app as they discover, but it should also continue to grow as search engine destination, especially in categories like home décor, food and beverage, and clothing.
- Brands will need to consider search intent within their Pinterest content strategy – from SKU database imagery to how-to inspiration content – and ensure that the descriptions within the Pins are reflective of how people may search for the product, as these annotations are key to producing relevant visual results.
- Brand should know their value and price-point comparison in the mark as this will increase competition for lookalike products.
- Consider offline to online opportunities, i.e. how your magazine spread can encourage visual searches.
How To Get Found in Image Search
While old-school, text-focused SEO tactics may be less relevant here, brands will need to understand searchers’ intents and desires (even keywords) more than ever. They’ll need to evolve their content strategy to recognize what consumers are seeking in visual content and how their brand’s value proposition can be best emphasized.
Pinterest’s search platform has evolved both visually and via categorization and affinities. After first analyzing what visual content your target customers want, image optimization will focus on a 3-pronged approach:
- High quality, easily discernable product images – intended for Flashlights results
- Engaging- intended to trigger image searches – Pinterest does best with evergreen content that isn’t dependent on quickly expiring products.
- Optimized descriptions, categorization & tagging – while the search will largely be based on the image itself, Pinterest will rely heavily on their own interest/affinity taxonomy and pin descriptions to make content suggestions.
Pinterest’s visual image search now allows Pinners to search for “all those things you don’t have the words to describe.” While purely an organic option for now, this has huge potential for ecommerce opportunities for advertisers once it expands to include buyable pins. It is also likely to either make Pinterest a search engine destination, drive Google, Bing, etc. to evolve their image search capabilities, or both!
For more information, visit the resources below: