Pinterest provided insight into how its algorithm works while warning of the risks of relying too much on user engagement. The platform stressed that over-reliance on engagement to rank content can lead to a negative user experience and proposed opt-out signals as a solution.
To encourage other companies to follow suit and help “build a more inspired Internet,” Pinterest published a new document titled “Field Guide to No-Engagement Signals.”
Why we care By having a better understanding of how Pinterest’s algorithm works, brands can identify which metrics they should prioritize in order to ensure more exposure for their content.
What are the signs of non-involvement? Disengagement signals are generated from two main sources:
In-App Surveys: Give users the opportunity to provide direct feedback about the platform. For example, Pinterest may conduct in-app surveys to collect user information. Independent assessments of content quality: usually generated from manual tagging.
In addition to balancing engagement signals in content ranking, non-engagement signals allow Pinterest to align with its values. For example, Pinterest’s commitment to inclusion is supported by signs of non-engagement. When users specify body type, hair pattern, or skin tone preferences in their feed, Pinterest can prioritize relevant content accordingly.
Field guide Pinterest teamed up with UC Berkeley and the Integrity Institute to create the Non-Engagement Signals Field Guide to help platforms create better user experiences over time. Pinterest noted that the goal of the guidance is simply to help platforms make informed decisions when it comes to using opt-out signals, rather than telling them what to do.
key points The field guide, which is based on practical industry knowledge, offers several practical applications for product development, including:
How to tune in for emotional well-being. Using generative AI to scale content quality signals. Improve user retention.
What Pinterest Says Pinterest Senior Data Scientist Leif Sigerson and Pinterest Senior Public Policy Officer Wendy Matheny said in a blog post:
“User engagement is a critical signal used by Pinterest and other online platforms to determine what content to show users. However, it’s widely known that optimizing solely for user engagement can show poor content quality (eg clickbait) or even harmful”. “Our CEO, Bill Ready, explained that if we’re not careful, content classification can get out of the ‘car crash we can’t take our eyes off of.'” On the other hand, if you ask someone later after seeing the accident, do you want to see another one?” the vast majority of people will say ‘God no'”. “Disengage signals are a critical component of making sure we don’t optimize for the ‘car crash we can’t look away from.’
Get the daily search newsletter marketers trust.
deep dive. Read the Pinterest one blog in full or download a PDF copy field guide for more information.
