PROTOCOL::CONTENT
How to Write Tweets for High Engagement
Understanding "high-scoring" tweet structures under multi-action prediction: from hooks to interaction points.
TL;DR // CORE
Executive Summary
- Use strong hooks to drive clicks and dwell time.
- Design open questions or "reply-able" points to spark interaction.
- Ensure content value to minimize negative feedback risks.
01 // LOGIC
Why It Matters to the Algorithm
X's recommendation model (Phoenix) doesn't just look at "Likes." It predicts the probability of multiple actions simultaneously. Among them, Replies, Reposts, and Dwell Time often carry significant weights.
If your tweet grabs a user in the first 500ms and makes them want to click or reply, the system marks it as high-value and expands its reach.
02 // SIGNALS
What the Algorithm Predicts
| Signal | Behavior | Tactic |
|---|---|---|
| P(Click) | Probability of clicking "Show more" | Use intriguing hooks; put the "meat" in the middle. |
| P(Reply) | Probability of a user replying | Ask relatable or controversial questions. |
| Dwell Time | Reading duration | Use threads or high-quality media. |
03 // CHECKLIST
Do & Don't Checklist
Should Do
- Niche Focus: Help the model categorize you.
- Interaction Hooks: End with a clear CTA.
- High-Quality Media: Boosts dwell time significantly.
Avoid
- External Links: Put them in replies instead.
- Clickbait Overload: Triggers negative flags.
- Spammy Posting: Diversity penalties apply.