For more than two years now, Instagram has used an algorithm that ensures that posts are no longer shown chronologically. There have been many rumors about how the algorithm works and how you can manipulate it. Finally, the time has come, and Instagram has disclosed the secret of how the algorithm works. Posts are evaluated taking various factors into account and then collected in a personalized form in each user’s feed. But what are these factors? We have summarized the most important points for you.

Interest – The posts are shown in the feed are selected based on other posts the user has often liked. For example, if a user often likes animal images, the likelihood of seeing more animal images in the future increases.

Timeliness – The Instagram algorithm is designed to show users the latest posts. New posts are shown first. However, this only works if the frequency a person uses Instagram is taken into account. To prevent users who are online several times a day from seeing the same content all the time, some older posts are ranked higher. If a user opens the app 20 times a day, the posts are only a few minutes old. If one opens the app only rarely, older posts are still at the top of the feed and are considered as current.

Relationship – This factor is similar to interest. As with most algorithms, it is influenced by previous interactions. If someone interacts with an account frequently, posts from this account will appear in the user’s feed more often. As a consequence, it should be clear that not only the number of followers is crucial to achieving a large reach – it is also important to have many interactions. However, instead of sorting out some posts, the content is only prioritized. You can still find all posts if you scroll down long enough.

Duration of use – The time span a person actually uses the app also affects which posts are shown in the feed. It makes a difference whether you open Instagram for a few seconds or for a longer period of time. If someone usually opens the app for a very short time, the user will be shown posts at the top that the algorithm considers most interesting. That implies that the factors explained above are more crucial.

Following – The number of profiles someone follows is crucial, too. The more accounts a user follows, the larger the number of images the algorithm can choose from. Additionally, the more profiles someone follows, the stronger interest, timeliness and relationship factors are taken into account by the algorithm to create the feed.

Stories – All these factors also determine the order in which stories are shown. For a long time, the Instagram algorithm was assumed to rank accounts higher in the feed that use the story function more frequently. But now we know that this is not the case. How often the story feature or the Instagram Live feature is used does not affect the composition of the feed.

Videos – Contrary to previous assumptions, videos are not ranked higher than pictures.

Company profiles – With Instagram explaining how its algorithm works, it also became clear that company profiles are not treated differently than individuals’ accounts. It is often claimed that companies that place advertisements get less reach. This is not correct either. The Instagram algorithm does not remember who spends money on posts and then ranks unsponsored posts lower to make companies pay again. Falling numbers in reach might be a result of the enormously grown competition in recent years instead of the algorithm discriminating unsponsored posts.

However, why did Instagram introduce the algorithm at all?

Before the algorithm was invented, 70 percent of users viewed only about 50 percent of the subscribed content in the feed. According to Instagram, the algorithm enables users to see 90 percent of the posts. Since the number of active users and the time spent on Instagram both increased, the invention of the algorithm can be considered a success. The number of active users per month exceeded 1 billion by June 2018, while it was 500 million two years before. These numbers support the necessity of the algorithm.

– For questions and suggestions please contact HBI