Instagram plans to significantly change its content recommendation algorithms, prioritizing original content and expanding the audience of smaller accounts.
About this informs The Verge with reference to the administration of the platform.
The main targets of these changes will be aggregator accounts that publish content created by other users, often without proper attribution to the authors. From now on, reposted content will be removed from recommendations on the platform. This will apply, in particular, to serial reposters – accounts that publish content that they “did not create and significantly improve” more than 10 times within a 30-day period.
In addition, Aggregators will also disappear from the Interesting page. These accounts will only be able to return to recommendations 30 days after the last posting of “unoriginal” content.
Also, the platform will replace content published by aggregators with original content in recommendations. This will be done if the original material is “relatively new” and the system is “confident” that the posts are identical “based on sounds and visual cues”.
It is important to note that the changes will affect only the recommendations. Users who are subscribed to aggregator accounts will still be able to see reposts in their profile and news feed.
In addition, Instagram will introduce a system of tags that will be added to reposts to indicate the original source of the content. Both the author and the reposter will be able to delete these tags. The new rules could have a significant impact not only on “content factories” that publish stolen memes and photos. Users who post infographics, star fan pages, or accounts that collect motivational quotes and photos from various sources may also be targeted.
The new recommendation system algorithm will also “give all authors an equal chance to break through.” Previously, the recommendation feed was formed based on the interaction of users with publications, that is, accounts with the largest number of followers often had a wider reach. Now the principle of operation of the system will be similar to beta testing: part of the content will be offered to a small audience that may be interested in it, regardless of subscription to the author. The most interesting posts will be presented to a larger group of users, and the process will repeat itself.
