Mastering the Twitter Algorithm: Main Points You Need to Know

In this article, we'll cover the main takeaways from the latest changes to the algorithm, including tips for boosting your visibility and avoiding demotion.

Mastering the Twitter Algorithm: Main Points You Need to Know

On March 31, 2023, Twitter open-sourced its algorithm, providing developers and researchers with access to the source code behind one of the most important parts of the Twitter experience.

The Twitter algorithm determines what users see on their "For You" page, notifications, and search results.

As soon as Twitter open-sourced 2 of its algorithms - the algorithm and the algorithm ml, developers around the world started to explore them and take some key points.

In this article, I've collected the most commonly found insights by now, and made a list of what should you do and what should you avoid for a better "twit", so let's dive in.

The Importance of Engagement and Interaction

Creating engaging content is crucial to getting noticed on Twitter.

The 'heavy-ranker' algorithm calculates various probabilities about user engagement with a tweet, including likes, retweets, replies, and clicks. User interaction also plays a major role in Twitter's algorithm.

Based on the source code, here're some interactions calculations:

ProbabilityWeight
The user will favorite the Tweet0.5
The user will click into the conversation of this Tweet and reply or Like a Tweet11*
The user will click into the conversation of this Tweet and stay there for at least 2 minutes11*
The user will react negatively (requesting "show less often" on the Tweet or author, block or mute the Tweet author)-74
The user opens the Tweet author profile and Likes or replies to a Tweet12
The user replies to the Tweet27
The user replies to the Tweet and this reply is engaged by the Tweet author75
The user will click Report Tweet-369
The user will ReTweet the Tweet1
The probability (for a video Tweet) that the user will watch at least half of the video0.005
Table of User Interaction Weights used by the "heavy ranker" model.
💡
'11*' is used to denote that the probability for the user to click into the conversation of a Tweet and reply or like a Tweet is weighted by a factor of 11 in the model.
The maximum prediction from these two 'good click' features is used, and the other prediction is ignored.

So, in the nutshell, for your tweet to accumulate more weight you should:

  • Encourage users to like or favorite your tweets.
  • Encourage users to retweet your tweet.
  • Encourage users to engage with your tweets by clicking on them and replying or liking them, or staying there for at least 2 minutes.
  • Encourage users to check out your profile and engage with your tweets there.
  • Create engaging tweets that encourage users to reply.
  • Create tweets that are likely to be engaged with by the tweet author.
  • Avoid eliciting negative feedback such as requests to show less often, blocking, or muting.
  • Avoid having users click "Report Tweet" on your tweets.

Another important key point of your twit is the quality of the accounts you interact with, it can influence your own visibility on the platform.

Twitter's Tweepcred algorithm, source code.

Interacting with low-quality accounts with bad follower/following ratios and being marked as spam/NSFW/bots/toxic can result in demotion.

Boosts and Demotions

Certain factors can give tweets a boost or result in demotion.

Here are some of the key takeaways:

Boosts:

  • Trends: Mentioning trending topics can result in a small boost.
  • Media: Tweets with images and videos can give your content a big 2x boost.
  • News: Tweets about current events or breaking news can receive a boost.
  • Verified accounts: Verified accounts get a big boost in the algorithm.

Demotions:

  • Offensive content: Posting offensive content can quickly demote your tweets.
  • Hashtags: Using too many hashtags can result in demotion.
  • No text, URL only, or name only: Tweets with no text or only a URL or name can also be demoted by the algorithm.

Twitter Blue and the Following-to-Follower Ratio

Twitter Blue subscribers get a boost in the algorithm, with a 4x boost if they're in the same network as the author of the tweet, and a 2x boost if they're not.

Twitter Blue boost algorithm, source code.

The following-to-follower ratio also matters.

Users with a low number of followers and a high number of followers may have their page rank reduced by the algorithm.

The Impact of Government Intervention

The government can intervene with the Twitter algorithm, particularly during important events like presidential elections.

Twitter engineers even have a class for government intervention, called GovernmentRequested.

During election events, the algorithm can recommend election candidates to follow and suppress misinformation.

Conclusion

In conclusion, here are some tips to optimize your tweets and avoid demotion by the Twitter algorithm:

Do:

  • Create engaging content that encourages user interaction like likes, retweets, and replies.
  • Use media like images and videos in your tweets to give your content a boost.
  • Mention trending topics in your tweets for a small boost.
  • Build relationships with high-quality accounts on Twitter to improve your own visibility.

Avoid:

  • Posting offensive content that can quickly demote your tweets.
  • Using too many hashtags in your tweets can result in demotion.
  • Creating tweets with no text or only a URL or name.
  • Interacting with low-quality accounts with bad follower/following ratios and marked as spam/NSFW/bots/toxic, which can result in demotion.
  • Following too many accounts compared to the number of accounts following you can result in deprioritization by the algorithm.
💡
This article is still in the writing process.

Follow me on Twitter or signup to this blog to read more about Angular and Generative AI!