Reach on X is not luck and it is not follower count. Since X open-sourced the core of its recommendation algorithm in March 2023, the broad shape of what the For You feed rewards has been public and, across repeated re-tunings, reasonably stable: it assembles a large candidate set of posts per session and scores each with a "Heavy Ranker" that weighs predicted engagement — and it does not value all engagement equally. In the open-sourced weights a reply is worth far more than a like (roughly 27x), and a reply the author then answers is the single most valuable interaction (around 150x a like), while reposts, bookmarks, dwell time, and profile clicks all feed the score and negative reactions — mutes, blocks, reports, "show less often" — cut it hard. On top of the weights sit two effects analysts see everywhere: a heavy early-engagement-velocity premium, where the first 30 to 60 minutes disproportionately set a post's reach, and a documented distribution edge for verified Premium accounts. This guide turns that machinery into strategy: what the algorithm is optimizing for, the real signal hierarchy and which signals are worth designing around, why the first hour matters so much, X's structural bias toward native on-platform content, where the contested link penalty actually stands after Elon Musk's July 2026 comments, and the durable posting playbook that follows — plus the operational reality that the playbook only pays off on repetition, which is where most creators stall.
Reach on X behaves like a system with rules, because it is one. Since X open-sourced the core of its recommendation algorithm in March 2023, the broad shape of what the For You feed rewards has been public, and across repeated re-tunings it has stayed reasonably stable. The feed pulls a large candidate set of posts per session and scores each with a model — the "Heavy Ranker" — that predicts how likely you are to engage, then ranks on a weighted blend of those predicted interactions. The single most useful thing to internalize is that the weights are not equal: the system values some actions dramatically more than others, and posting well means feeding it the ones it pays for.
This guide turns that machinery into a strategy you can act on. It covers what the algorithm is actually optimizing for, the real signal hierarchy and which signals are worth designing around, why the first hour after posting matters so much, X's structural bias toward native on-platform content, where the contested link penalty stands after Elon Musk's July 2026 comments, and the durable posting playbook all of that dictates. One honesty note up front, because it governs everything below: the specific multipliers come from the 2023 open-sourced code, and X has re-tuned its system since without publishing every current weight — so treat exact numbers as directional and build on the ordering, which has held. For the definitional groundwork, see algorithm and engagement rate.
Strip away the folklore and X's ranking is a prediction problem. For each session it gathers candidate posts — from accounts you follow and, increasingly, from outside your network — and the Heavy Ranker estimates the probability you will reply, repost, like, bookmark, click the profile, or dwell on each one. It then combines those predictions into a score using fixed weights, and the highest-scoring posts fill your feed. Follower count is not a direct input the way people assume; what matters is predicted engagement and the reputation of the author, which is why a small account posting reply-worthy content into an active window can out-distribute a large, passive one.
The consequence for strategy is that you are not writing for humans and an algorithm as two separate audiences — you are writing for humans in a way that produces the specific signals the model is trying to predict. A post that earns a genuine reply, a save, or a few seconds of real attention is a post the ranker scores highly, because those are exactly the outcomes it was built to forecast. That reframes the whole exercise: the job is not to trick the system but to reliably produce the behavior it rewards. Everything in the rest of this guide is a more specific version of that one idea.
The most important thing the open-sourced weights revealed is the ordering. A like is a cheap, passive tap, and it is weighted accordingly — as the baseline. A reply is worth far more: in the released code, roughly 27 times a like. And the top of the hierarchy is a reply that the author then engages with — a back-and-forth exchange — weighted around 150 times a like, because a real conversation is the costliest, most meaningful signal a post can generate. Reposts sit above likes as a stronger endorsement, and profile clicks count as a sign the post made someone want to know more about you. The blunt takeaway: the algorithm pays the most for conversation and the least for passive approval.
That single fact should reorganize how most accounts write. If your posts are engineered to collect likes — safe, agreeable, nothing to respond to — you are optimizing for the weakest signal in the system while leaving the strongest untouched. The higher-leverage move is to write posts that invite a response: a genuine question, a specific claim people will want to push back on or add to, a prompt to share their own version. And when the replies come, answer them — an author-engaged reply is the top-weighted interaction, and your responses restart the conversation that keeps a post climbing. This is the same logic that makes reciprocal relationships so valuable under X's mutual-interactions boost, covered in X now boosts mutual interactions.
Beneath the headline weights are a few signals worth designing around. Bookmarks are a strong quality signal — a save says "I want to come back to this," which is a stronger statement of value than a like and something the system reads as high intent. Practical, reference-worthy posts (frameworks, checklists, resource lists, genuinely useful threads) earn saves that a throwaway hot take never will. Dwell time — how long people linger — is another heavily weighted input, and it is won or lost in the first line or frame, so a sharp hook is not a stylistic nicety but a ranking lever. Video read-through works the same way: the longer people watch, the stronger the signal.
On the other side of the ledger are the negative signals, and they are not gentle. Mutes, blocks, reports, and "show less often" tell the ranker to suppress a post and can damage an account's standing over time. That makes engagement bait a bad trade: "like if you agree," follow-for-follow chains, and misleading hooks that make people feel tricked into a click all raise the odds of a negative reaction, and a reader who feels baited is one tap from muting you. Earning interaction honestly is not just the ethical path; it is the one that avoids the penalties that quietly cap everything else.
The weights describe how much each interaction counts. Velocity describes how the timing of those interactions is read, and it is one of the most consequential effects on the platform. X's ranking rewards speed of engagement, not just volume: a post that earns interaction quickly is amplified to a far wider audience than an identical post that earns the same interaction slowly. Analysts consistently find the first 30 to 60 minutes disproportionately determine a post's eventual reach — early engagement is the signal that tells the ranker this post is worth showing to people beyond your immediate network.
This turns two things you might treat as afterthoughts into real levers. The first is timing: publishing into the windows when your audience is actually online — which your analytics will show — gives a post the early engagement it needs to catch, while the same post dropped into a dead window can stall no matter how good it is. The second is presence: being available for the first hour to answer early replies compounds both the velocity effect and the author-engaged-reply weighting at once. The practical rule is to post when your people are awake and stay for the window, not to fire and forget. For structuring that into a repeatable rhythm across platforms, see social media posting schedules in 2026.
A pattern runs through every signal above: they all happen on-platform. Replies, reposts, bookmarks, dwell, and read-through are all interactions with a post while the reader stays on X. That gives the system a structural bias toward content that keeps people there, and it shows up in what performs. Native video tends to do well, particularly shorter pieces that earn full read-through. Polls hold attention and add a dwell signal almost by design. Images and threads keep people in the post longer than a bare line of text. None of this means plain text can't land — it often does — but native media hands the ranker more of the signals it is looking for.
Threads deserve a specific mention because they are the cleanest way to manufacture dwell. A well-built thread earns read-through as people move post to post, stacking dwell time, and every post in it is another surface that can collect replies and bookmarks. The craft is to make the first post hook completely and each following post pay off the last, rather than padding to hit a number — a tight three-to-five-post thread on one idea beats both a single cramped post and a bloated ten-post one. The through-line for format choice is simple: prefer whatever keeps the reader engaged on the post over whatever tries to move them off it.
Which brings up the most-argued tactic on X: links. For years the working assumption was that an external URL in the body of a post cut its reach, and the standard fix — post the content, drop the link in a reply — became near-universal advice. In 2026 that assumption got publicly challenged. On July 29, 2026, responding to Paul Graham asking whether X still throttles link posts, Elon Musk replied, "We haven't for over a year." Taken at face value, that would retire the workaround entirely.
The evidence, though, has not caught up to the statement. PPC Land's January 2026 technical analysis of linked versus unlinked posts reported independent testing showing link-in-body posts receiving dramatically fewer views — on the order of 94% fewer — and researchers responding to Musk in the same thread noted that indirect suppression could persist even without an explicit penalty flag in the code. So the honest status is unsettled: leadership says there is no penalty; independent measurement and a lot of creator experience still show link posts underperforming. When the downside of being wrong is a large reach hit and the cost of caution is trivial, the low-risk default holds — post natively and put the link in the first reply until the data agrees with the statement.
Two account-level factors sit underneath every individual post. The first is verification: X's ranking treats a verified Premium subscription as a credibility signal, and verified accounts have a documented distribution advantage over non-subscribers. That does not mean a subscription buys reach on its own — a Premium account posting weak, reply-less content still underperforms — but it removes a handicap, and for an account serious about growth on X it is part of the table stakes rather than a growth hack. The second is reputation: the system factors in an author's track record, so consistent, non-negative posting builds standing that lifts your baseline, while a run of low-engagement or bait posts erodes it.
Both point to the same discipline. Reach on X is not only about the individual post; it is about the account behind it, accumulated over time. That is why volume without quality is a bad trade — flooding the feed with mediocre posts to "stay active" can lower the very reputation that determines how far your good posts travel. A sustainable cadence of strong posts compounds; an erratic burst of weak ones works against you. The account is the asset, and every post either adds to or subtracts from its standing with the ranker.
Put the pieces together and the strategy is concrete. Post into your audience's active windows so posts earn fast early engagement. Open with a hook that wins the first two seconds and the dwell that follows. Write to start replies, not to collect likes, and be present in the first hour to answer them — the richest signal in the system is a conversation you are part of. Keep the content native — video, images, polls, threads — so the attention stays on X, and keep links in a reply. Hold a consistent, sustainable cadence rather than chasing volume, and avoid the bait that earns negative reactions. None of it is a trick; each item is just a more specific way of producing the behavior the ranker was built to reward.
Two framing points keep the playbook durable. First, build on the ordering, not the exact weights: replies over reposts over likes, fast over slow, native over outbound, and honest over baited have survived every re-tuning because they reflect what the platform is fundamentally trying to surface — bet on those and the next weight change tends to work in your favor rather than resetting you. Second, reach on X has tilted toward the tighter graph. X's 2026 changes, including the boost to posts from your mutuals, reward a reciprocal, active community over farming strangers, so the strongest position is a defined audience that consistently engages with you plus the occasional wide-reach post riding on top. The reciprocal-graph dimension is worked in full in X now boosts mutual interactions, and the same durable-signals logic underpins how ranking works on Instagram.
Here is the catch every honest strategy piece has to admit: the playbook above is not hard to understand and not hard to do once. It is hard to do repeatedly. Reach on X comes from posting into the right window, in native formats the ranker favors, at a sustainable three-to-five-a-day cadence, written to start replies, day after day after day — and that repetition is exactly where the tactics collide with the hours in a week. Most accounts stall not because they don't know the strategy but because producing that volume of good, native, on-brand posts by hand is a full-time job on its own, and the part that actually earns reach — being present to reply — is the part that has to stay human.
That split is where Kompozy earns its place. It is a full AI content generation and multi-platform publishing engine — 18 output formats, not a repurposing add-on — and its job here is to carry the production so your attention goes to the conversation. Working from one Persona Brief that fixes your voice, recurring points of view, and banned words, it generates the native X format mix the ranker rewards: Text Posts built to open with a hook and end on a reply-worthy prompt, Persona Tweets that pair a face-locked image with a tweet-card composite for the native-media edge, threads for dwell, and short Persona Shorts video for the moments a talking-head clip lands harder than text — all in one recognizable identity, with HyperFrames keeping the styling brand-exact.
The scheduler and Autopilot then let you queue that cadence into your audience's active windows — the early-velocity effect the whole strategy turns on — behind a per-post review gate so you approve every post before it ships, and you can line the link up as the follow-up reply the workaround calls for. Because it is a real generation engine, the same identity extends to every other platform your audience lives on, so X becomes one native surface in a coherent presence rather than a separate manual chore. What Kompozy deliberately does not do is reply to people for you — the reply-weighting reward is human conversation, and automating it is precisely the drive-by behavior the system demotes. It keeps you consistently, natively present so you are free to spend your actual attention on the exchanges the algorithm pays for. That is the right division of labor for an X that ranks conversation first: let the engine produce the presence, and keep the relationships human.
X algorithm posting strategy in 2026 is not a set of hacks; it is a short list of durable truths about what the ranking system rewards. It weighs conversation over passive approval — replies far above likes, an author-answered reply above all — reads engagement velocity so the first hour disproportionately sets reach, favors native content that keeps attention on the platform, treats the link question as unsettled enough to warrant caution, and factors in the reputation of the account behind every post. Build on that ordering rather than on any exact multiplier, hold a sustainable native cadence into your audience's active windows, and spend your human attention on the replies. The strategy is knowable; the winning move is doing it consistently, which is a production problem before it is a knowledge one.
X's For You feed assembles a large candidate set of posts per session and scores each with a "Heavy Ranker" model that weighs predicted engagement — and not equally. From the algorithm X open-sourced in March 2023, a reply is weighted far more than a like (roughly 27x) and an author-answered reply is the top signal (around 150x a like); reposts, bookmarks, dwell time, and profile clicks also count, while negative reactions cut reach. Two effects sit on top: a strong early-engagement-velocity premium in the first 30–60 minutes, and a distribution edge for verified Premium accounts. Treat exact numbers as directional; the ordering is the durable part.
Conversation. In X's open-sourced weights a reply is worth about 27x a like, and a reply the author then engages with is the single most valuable interaction — around 150x a like. That makes posts written to start a discussion (a real question, a take worth arguing with, a thread that invites responses) out-distribute posts written to collect passive likes. Reposts and bookmarks matter too, but the clear hierarchy is replies over reposts over likes, so the highest-leverage change most accounts can make is to write for replies instead of likes.
It's disputed. On July 29, 2026, Elon Musk said X had not penalized link posts "for over a year," but a January 2026 PPC Land analysis reported testing showing link-in-body posts receiving roughly 94% fewer views, and researchers note indirect suppression could persist without an explicit penalty. Because the evidence hasn't caught up to the statement, the low-risk practice is unchanged: post your content natively and put the URL in the first reply. You keep the on-platform reach the algorithm rewards and still route interested readers where you want them.
X rewards engagement velocity — how fast a post earns interaction, not just how much. Analysts consistently find the first 30 to 60 minutes disproportionately determine a post's eventual reach, because fast early engagement signals the ranker to amplify the post to a far wider audience than an identical post that earns the same engagement slowly. That makes two things real distribution levers: posting when your audience is actually online, and being present during that window to answer early replies, which restarts the conversation the algorithm pays for.
Consistency builds the account reputation the ranker factors in, but volume past your quality ceiling backfires — each low-engagement post can drag your account's standing and lower the reach of the next one. Many creators land on a sustainable rhythm of roughly three to five strong posts a day rather than flooding the feed. The goal is a cadence you can hold at quality, because posting steadily keeps you visible enough to earn the replies and reposts the rest of the strategy depends on.
It's a weaker single bet than it was. The For You feed still surfaces strong posts beyond your network, so a genuinely good post can travel — but X's 2026 tweaks (including a boost to posts from your mutuals) have tilted the payoff toward a tighter, active audience graph over farming strangers. The durable strategy is to build a reciprocal community and a consistent native presence, then let the occasional wide-reach hit ride on top, rather than betting everything on virality. See the companion guide on X boosting mutual interactions for that dimension.
X algorithm posting strategies for 2026 come down to what the ranking system rewards. Its Heavy Ranker weighs predicted engagement unequally: replies count far more than likes (about 27x in the open-sourced code), an author-answered reply most of all, and reposts, bookmarks, and dwell time feed the score. Early engagement in the first 30–60 minutes disproportionately sets reach. So post into your audience's active window, write to start replies, keep content native, and put links in a reply — building on the durable ordering, not exact weights.
Get started → · ← All guides · Compare Kompozy vs other tools