- Pinterest revealed insight via its engineering blog last week into its evolving "Pinnability" algorithm.
- Engineers at Pinterest are utilizing machine learning models to deliver the Pins it determines are most relevant to users.
- The algorithm assigns a personal relevance score to each Pin eligible to appear on a user's feed—a mix of Pins from users they follow and "Picked for You" Pins.
Pinterest faces the problem of how to deliver relevant content amid an overwhelming amount of Pins. The site hopes the algorithm will serve a dual purpose: delivering a better user experience and offering better targeting for advertisers.