How Twitter's For You Algorithm Distributes Views
Most authors on Twitter face the same problem: their posts disappear into the void. Good text, correct timing, a trending topic — but no views. The reason is almost always the same: the post didn't make it into the For You feed. Twitter's For You algorithm is not a random selection of content for recommendations. It's a multi-level evaluation system that decides who sees a post beyond the author's followers. Understanding how it works means you stop publishing blindly and start working with the mechanics that truly drive reach.
What is the For You feed and how does it differ from the Following feed?
The Twitter For You feed is an algorithmically generated section of the homepage that shows users posts regardless of whether they follow the author. This is a fundamental difference from the Following tab, which only displays content from accounts the user follows.
Getting into For You means getting impressions from an audience that has never heard of the account. This is where organic growth happens: new users see the post, interact with it, and subscribe. For most authors, the Twitter recommendation feed is the main source of new views and followers.
X released an updated For You algorithm on GitHub in May 2026, giving developers and marketers access to the real ranking logic. This is one of the few cases where a major platform openly publishes the principles of its recommendation system — and it's worth taking advantage of.
How the For You algorithm evaluates posts for recommendations
Ranking X.com posts in the For You feed goes through several stages. First, the system selects candidates from a huge array of publications, then ranks them by a set of signals, and only then shows the user a personalized selection.
At the initial selection stage, the algorithm looks at thematic relevance: the user's interests, their interaction history, and the accounts they follow. Posts that do not match the user's interesting cluster are filtered out even before ranking.
At the ranking stage, the Phoenix model is activated — a system for evaluating predictable user actions with a post. The algorithm calculates the probability that the user will like, retweet, comment, or click "not interested." Posts with high predictable engagement are prioritized.
An additional filter is account reputation. The X platform uses the TweepCred metric: an internal assessment of an account's authority and activity. Accounts with high TweepCred get an advantage at the start — their posts are more likely to enter the primary pool of candidates for recommendations.
Post views as a signal for the X algorithm
The X algorithm perceives post views not as an independent indicator, but as an input signal for evaluating behavioral patterns. The number of views itself does not rank a post higher — what happens during these views is important.
Dwell time has become one of the key ranking factors in 2025–2026. The algorithm tracks how long a user spent on a post before scrolling further. A post that is read longer receives a high-value signal — and the algorithm expands its reach in the For You feed.
This is why views combined with high dwell time have a fundamentally different effect than views with quick scrolling. A post that is opened, read, and reacted to accumulates algorithmic weight significantly faster than a post with the same number of views but without attention delay.
Engagement signals and reach in the For You feed
Engagement and Twitter reach are linked through algorithmic logic: each type of interaction with a post carries a different weight when ranking in the For You feed.
Comments are the strongest signal. The algorithm interprets them as a sign that the post provoked a reaction strong enough for the user to write a reply. Posts with active discussions in the comments receive extended reach — the algorithm assumes that if some users engaged in discussion, others will want to too.
Quotes work similarly to comments, but with an additional effect: each quote creates a new post that can also get into For You. This is a reach multiplier — one original post generates a chain of posts in different audience segments.
Retweets directly expand reach by showing the post to the followers of the person who reposted it. The algorithm also uses retweets as a signal "the user considers this valuable for their audience" — and increases the post's priority.
Likes are the least weighty signal among those listed, but important for an initial boost. In the early stages of dissemination, likes from active accounts trigger the first stage of reach expansion and give the post a chance to enter the primary For You pool.
Saves are a relatively new signal that the X algorithm considers as an indicator of long-term content value. Posts with a high number of saves remain active in the distribution system longer.
How the Twitter recommendation feed works: practical factors
Understanding how the Twitter recommendation feed works allows you to build a content strategy based on specific algorithmic signals, rather than acting intuitively.
Relevance and freshness. The For You algorithm prioritizes fresh content — especially content related to actively discussed topics. Posts that appear at the beginning of a trend receive a boost from thematic clustering and reach an audience that has already interacted with similar content.
Our Services for Streamers
Our Services for Content Creators
