Pinterest is a platform designed around image sharing and social media feedback (i.e. likes, comments, shares) using a pinboard metaphor in both its design and function. Coming into its initial popularity in 2012 (Madrigal, 2014), Pinterest allows users to curate collections of web content represented through images. One of its unique features is an accompanying plug-in that allows users to capture data on other sites in order to turn them into “pins” or posts for Pinterest that link to the original content; as a result, Pinterest users enter new visual data into Pinterest even as they build out their own collection. Pinterest functions as a database “where people can go to get ideas for any project or interest in their life,” according to co-founder Evan Sharp (Madrigal, 2014). Pinterest has fairly extensive community guidelines that prohibit against particular kinds of content, including violent content and misinformation that Pinterest “limit[s] the distribution of or remove[s] such content and accounts.”
Before you Search
Before a user navigates to the search bar, if they are logged in, they see their home feed (see above). While it isn’t a search function, per se, the home feed pushes content that is linked to the user’s previous clicks or pins in some way. Recently, Pinterest added an option to “edit” or “tune” one’s home feed, which suggests content based on a user’s browsing history, boards they follow, topics they follow or are popular, and profiles they are linked to.
How to Search
As of April 2022, Pinterest users can both informally browse the site and search using the main tool bar at the top of the page on the desktop version. On the mobile version, the search icon is located at the bottom of the application. In addition to suggestions that show up visually when the user clicks the search bar, users can also search their own pins or all pins using key terms.
Desktop version
Mobile version
How does Search work?
Because Pinterest is a largely visual site (even in terms of how one experiences the site), its algorithm is designed to acknowledge the visual: essentially, it is a visual search engine that users navigate primarily through images, rather than key words. (However, it is important to note that though users tend to interact with content based on visual cues, all posts are linked to key words that function as part of the algorithm.)
When a user searches for content on Pinterest, their individual results vary according to several factors including what other users they they follow and content they have pinned to their own boards that might be related to the search query (DeLuca, 2012). And though we don’t know exactly how Pinterest’s algorithm works because it is a proprietary secret, Pinterest users have determined that there are four main ranking factors for the algorithm–that is, what pins get shown to users first (Mikke, 2021).
Domain quality: how popular is the website linked through the pin and how much content from the site is on Pinterest?
Pin quality: how popular and timely is the pin itself?
Pinner quality: how popular and engaged is the user that created the pin?
Topic relevance: how popular are the Pinterest keywords linked to the pin?
Each of these factors is ultimately related to engagement. Pinterest calculates engagement through clicks (clicks on content that leads to a site outside of Pinterest); long clicks (clicks that result in 30 seconds or more of engagement); pin clicks (clicks to zoom into a pin from a board); and saves (saving a pin to a personal board). Certain types of pins have been introduced to encourage greater engagement, including Idea Pins which function as a sort of Instagram Story (Miranda, 2022).
There are some other features of Pinterest’s algorithm that shape search results, as well. One of those is “controllable distribution,”’ which ensures that users are greeted with content they would otherwise most likely not encounter. First, Pinterest returns results using their traditional algorithm that takes into account the four factors above; then, the controllable distribution algorithm assesses the variety of the results and makes changes based on those results to bring users different results (Grief, 2020).
“Pin freshness” is another factor, which simply refers to how closely a pin is related to other content. A fresh pin is one that has never been used on Pinterest before; specifically, an image that has never been used before. The fresher the pin, the better it might fare in the algorithm (Matthews, 2020).
Given Pinterest’s status as a visual search engine, they incorporated the use of “Pinterest Lens” in 2017. Lens is a features similar to Google Lens, as it analyzes a picture uploaded by a user and suggests search results that are similar to that image. In 2020, Pinterest linked this feature to their “Shop” tab to increase commerce (Hanz, 2021).
In February 2022, Pinterest began to promote a special type of pin called an “Idea Pin,” which is essentially a Pinterest version of an Instagram reel, using video to engage users. Because these are more work to create, they are often used by businesses or influencers on Pinterest, resulting in more focus on their content than everyday users (Anastasia, 2022).
Takeaways
Because Pinterest is a visual search engine, there are possible copyright and privacy concerns that accompany content.
Pinterest’s design that allows and even encourages users to link new content by “pinning” content from external sources represents an opportunity for users to generate data for Pinterest, meaning that users are an important node for search functions as they can create tags, etc. for content.
User engagement seems to be the primary metric that shapes the search algorithm for Pinterest, suggesting that the results pushed to users depend heavily on high levels of digital circulation.