The latest articles on DEV Community by Golam Kabir (@iamgolamkabir). https://dev.to/iamgolamkabir https://media2.dev.to/dynamic/image/width=90,height=90,fit=cover,gravity=auto,format=auto/https:%2F%2Fdev-to-uploads.s3.us-east-2.amazonaws.com%2Fuploads%2Fuser%2Fprofile_image%2F4158791%2Fd28a53cf-5070-4146-a0af-b19c51ec6eeb.jpg https://dev.to/iamgolamkabir en Golam Kabir Sat, 03 Oct 2026 02:54:41 +0000 https://dev.to/iamgolamkabir/x-engagement-rate-formula-and-benchmarks-1j97 https://dev.to/iamgolamkabir/x-engagement-rate-formula-and-benchmarks-1j97 <h2> Key takeaways </h2> <ul> <li>The X engagement rate formula on this site is total interactions divided by impressions, times 100. Interactions here are likes, reposts, replies, bookmarks, and clicks.</li> <li>X's own analytics definition has changed before. Treat this formula as the one the calculator uses, not as a permanent official standard.</li> <li>The ranges below are rules of thumb I use when reading an account. They are not a published study, and they are not X's targets.</li> <li>A higher rate is not automatically better. A tiny, irrelevant audience can post a flattering percentage.</li> </ul> <p>An X engagement rate is a percentage that answers a narrow question: of the times this post was shown, how often did someone do something other than scroll past?</p> <p>It does not answer whether those people can buy. It does not answer whether the ranker will show the post again. It does not prove the writing was good. I still use it, because a rate you can recompute is better than a vibe, as long as you remember what was left out.</p> <p>I have looked at this number across 500+ accounts and 10+ years of X work. I am not going to pretend that history is a published benchmark study. It is not. The ranges later in this article are rules of thumb, labeled as such. The arithmetic is the part you can check.</p> <p>The <a href="https://golamkabir.com/tools/x-engagement-calculator" rel="noopener noreferrer">X engagement rate calculator</a> does the arithmetic in the browser. This page explains the formula it uses, the other formula people confuse it with, and how not to lie to yourself with either one.</p> <h2> The formula this site uses </h2> <p>For a single post:</p> <p><strong>Engagement rate = (likes + reposts + replies + bookmarks + clicks) / impressions × 100</strong></p> <p>For a 30-day account view, the calculator uses the same formula on 30-day totals: all of those interactions over the period, divided by all impressions over the period, times 100. It is an average weighted by impressions, which is what you want. It is not the average of each post’s rate, which lets a tiny post swing the story.</p> <p>A worked example, using round numbers so the division is obvious. These numbers are an illustration, not a result from a client.</p> <ul> <li>10,000 impressions</li> <li>200 likes</li> <li>40 reposts</li> <li>30 replies</li> <li>20 bookmarks</li> <li>10 link or detail clicks</li> </ul> <p>Interactions = 200 + 40 + 30 + 20 + 10 = 300.</p> <p>300 / 10,000 = 0.03.</p> <p>0.03 × 100 = 3%.</p> <p>If you leave clicks out, you get 290 / 10,000 = 2.9%. Say which version you used. I include clicks because a click is a deliberate action and because the calculator includes a clicks field. If your export from X does not list clicks, leave the field at zero and compare only against other posts you measured the same way. Do not invent the missing clicks.</p> <p>Quotes are not a separate field in the calculator. If your analytics bundle quotes with reposts, they are already inside the repost count you type. If they are separate and you care, add them into the repost field and be consistent next time. Consistency beats a perfect taxonomy you cannot repeat.</p> <h2> Why this is not “the” official formula </h2> <p>X’s analytics product has described engagement in its own way, and that description has not stayed frozen. I will not crown one in-product tooltip as the permanent law. If you are reporting to a team, write the formula in the footnote. “Engagement was 2%” means nothing until the reader knows the numerator and the denominator.</p> <p>Two common variants:</p> <p><strong>Impressions in the denominator.</strong> This is the formula above. It asks how interesting the post was to the people who actually saw it.</p> <p><strong>Followers in the denominator.</strong> Some older social reports divide interactions by follower count. That made slightly more sense when reach and followers moved together. On X they often do not. A post can be shown to far more people than follow you, or to far fewer. Dividing by followers then flatters a post nobody saw (a few likes from friends, a huge rate) and punishes a post that traveled (many interactions, divided by a follower count that did not grow in the same hour).</p> <p>I use impressions. If a tool you bought uses followers, do not paste its percentage next to mine and call the gap a trend.</p> <p><strong>A per-post average versus a totals rate.</strong> If you average the rates of ten posts, a post with 200 impressions and 20 likes (10%) can outweigh a post with 100,000 impressions and 1,000 likes (1%) in a simple mean. The totals method does the opposite: it lets the large post dominate, which matches how much attention you actually received. The calculator’s 30-day mode is the totals method. Use it when you are judging the month. Use single-post mode when you are judging a draft’s siblings.</p> <h2> What each term is allowed to mean </h2> <p>Likes are cheap. They belong in the rate because they happened. They are a weak commercial signal.</p> <p>Reposts mean someone was willing to put the post on their own profile. That is stronger than a like, and still not a lead.</p> <p>Replies mean someone wrote back. Inside this formula they count as one interaction, the same as a like. Inside X’s published For You weights they do not. As of the 26 Sep 2026 check of the public file, a reply is weighted 5.0 and a like is 0.5, on predicted probability, not on raw counts. I explain that properly in <a href="https://golamkabir.com/blog/how-the-x-algorithm-works-in-2026" rel="noopener noreferrer">how the X algorithm works in 2026</a>. Read the rate, then read the mix. A 2% rate that is almost all likes is a different post from a 2% rate with a real reply thread.</p> <p>Bookmarks mean someone saved the post. The public For You file this site checked does not list a bookmark weight. I still count bookmarks in the rate, because a save is evidence the post was useful to a person. Do not convert that choice into “X ranks bookmarks at N.” I do not have N.</p> <p>Clicks, in the calculator, are whatever click count you decide to paste: link clicks, profile clicks, or a column your export already summed. Label it. A click to your site is not the same event as a click to expand a photo, but both are more deliberate than a scroll.</p> <p>Impressions are shows, not people. One person can see a post more than once. A rate can move because the denominator changed, not because the writing changed. If impressions collapse and the rate rises, you may have a smaller, warmer pocket of viewers. That can be good or bad. Look at who they are.</p> <h2> A good engagement rate on X, as a rule of thumb </h2> <p>This is the section that usually gets copied without the warning. Here is the warning, in the same size type as the ranges.</p> <p><strong>These bands are rules of thumb used by the calculator on this website. They are not an industry study. They are not medians from a dataset I am willing to cite. They are not targets published by X. They are labels I find useful when a founder asks “is this post asleep?” Do not put them in a deck as research.</strong></p> <p>For a <strong>single post</strong>, the calculator’s labels are:</p> <ul> <li>Under 0.8%: look at the first line, the audience, and whether anyone had a reason to reply. The post may still be fine if the impressions were broad and accidental.</li> <li>0.8% to 1.8%: a working baseline on this page. Not a prize. Not a failure.</li> <li>Above 1.8% and up to 3.5%: stronger than that baseline. Check that the extra interactions are from people you wanted.</li> <li>Above 3.5%: unusual relative to these labels. Before you call it a win, check whether the impression count was small. A high rate on a tiny audience is easy to celebrate and hard to repeat.</li> </ul> <p>For a <strong>30-day account total</strong>, the calculator uses different labels, because a month of posts will not look like your single best post:</p> <ul> <li>Under 0.6%: the month is quiet. Look at reply habits and whether most posts are links with no conversation.</li> <li>0.6% to 1.4%: a working month on this page.</li> <li>Above 1.4% and up to 2.8%: stronger than that working month.</li> <li>Above 2.8%: unusual for a broad month of impressions. Again, check the denominator before you frame a trophy.</li> </ul> <p>I changed the wording on the calculator so it does not call these an industry median or a top percentile. Older pages on the internet will still do that. They are guessing, or they are citing a report that measured a different formula on a different year. If you need a comparison, compare your March to your April on the same formula. That is a benchmark you actually own.</p> <p>Account size changes the feel of a rate. A small, tight audience often shows a higher percentage than a large, mixed one, because more of the viewers already care. I am not going to publish a table of “under 10k followers should be X percent.” I do not have a study behind that table. Use the bands above as a first glance, then segment by whether the viewers match the offer.</p> <h2> When a flattering rate is a bad sign </h2> <p>A good engagement rate on X can still be a bad account.</p> <p>The audience is wrong. Friends and peers will reply to you. Buyers might not be in the room. The rate looks healthy. The pipeline does not move. Spot-check who replied.</p> <p>The post traveled for the wrong reason. Outrage, a celebrity mention, or a joke that has nothing to do with the work can spike interactions. You will feel the number. You will not feel the business. I would rather see a duller rate from operators than a brighter rate from a crowd that will never hire you.</p> <p>The denominator is tiny. Twenty impressions and four likes is 20%. It is also almost no information. Wait until a post has been shown enough times that a couple of extra likes cannot swing the percentage wildly. I will not pretend there is a magic impression floor. Use judgment. If you would not bet on the rate surviving another hundred viewers, do not report it as a result.</p> <p>The numerator is padded. Pods, engagement groups, and bought replies raise the rate and teach you nothing. They also sit next to the negative actions in X’s published ranking file. I do not use them, and I do not count them as skill.</p> <h2> How to raise the rate without gaming it </h2> <p>Start with the first line. If a stranger cannot tell what the post is about, they will not interact. The <a href="https://golamkabir.com/tools/x-hook-scorer" rel="noopener noreferrer">hook scorer</a> is a checklist for that line, not a virality oracle.</p> <p>Ask for a reply only when the reply is the point. A real question gets fewer responses than a bait question, and the responses are the ones you can use. Remember that replies and likes are equal in the rate and unequal in the published ranker. If you care about distribution, a conversation beats a like-farm. If you care about the rate alone, you can inflate it with easy likes and learn the wrong lesson. Care about both, and read them separately.</p> <p>Make something worth saving if you want bookmarks. A framework, a checklist, a specific teardown. “Bookmark this” as a command is weaker than a post that is obviously worth keeping.</p> <p>Move the link on a test post. I do not have a published penalty percentage. I do have enough practical experience to test one post with the link in the body and one with the link in a reply. Compare rates and, more importantly, compare the replies. One test is not a law.</p> <p>Reply under other people’s posts if your own impressions are stuck. Engagement rate on a post nobody sees will not save the account. Distribution comes first. The map is in <a href="https://golamkabir.com/blog/how-to-find-the-right-people-to-engage-with-on-x" rel="noopener noreferrer">how to find the right people to engage with</a>, and the wider system is in <a href="https://golamkabir.com/blog/how-to-grow-on-x-in-2026" rel="noopener noreferrer">how to grow on X in 2026</a>.</p> <h2> How to use the calculator without fooling yourself </h2> <p>Open the <a href="https://golamkabir.com/tools/x-engagement-calculator" rel="noopener noreferrer">engagement rate calculator</a>. Pick single post or 30-day. Type the impressions and the interactions you actually have. Read the percentage. Read the label as a rule of thumb. Then write one sentence in your notes: what was in the numerator, what was in the denominator, and whether the people who interacted are the people you want.</p> <p>Do that for five posts before you change your whole strategy. One post is an anecdote. Five posts is a pattern you can argue with.</p> <p>If the pattern is “links die, replies live, the bio converts,” you do not need a more complicated metric. You need <a href="https://golamkabir.com/services/x-engagement" rel="noopener noreferrer">X engagement strategy</a> if the replies are in the wrong rooms, or <a href="https://golamkabir.com/services/x-organic-growth" rel="noopener noreferrer">organic growth</a> if the whole loop is missing. The calculator will not make that call. It will stop you from arguing about a feeling.</p> <p>That is all an engagement rate is for. A shared formula, a labeled rule of thumb, and the discipline not to promote either one into a study.</p> <p>SEARCH &amp; AI FAQ</p> <h2> Frequently Asked Questions </h2> <p>Direct practitioner answers to common industry inquiries.</p> <p>What is the X engagement rate formula?</p> <p>On this site, engagement rate equals (likes + reposts + replies + bookmarks + clicks) divided by impressions, multiplied by 100. The free calculator at golamkabir.com/tools/x-engagement-calculator uses that formula for a single post or a 30-day total.</p> <p>What is a good engagement rate on X?</p> <p>There is no official good rate. As a rule of thumb inside this site's calculator, a single post under 0.8% is a prompt to check the hook, 0.8% to 1.8% is a working baseline, and above that is stronger than those labels. These bands are not a study.</p> <p>Should engagement rate use followers or impressions?</p> <p>Use impressions when you have them. Dividing by followers punishes posts that reached beyond your following and flatters posts that almost nobody saw. Compare one post with another on the same account before you compare yourself with a stranger.</p> <p>Do likes and replies count the same?</p> <p>They count the same inside this engagement-rate formula, because both are interactions. They do not count the same in X's published For You weights, where a reply is weighted far more heavily than a like. Read the rate, then look at the mix.</p> <p><em>Originally published at <a href="https://golamkabir.com/blog/x-engagement-rate-formula-and-benchmarks" rel="noopener noreferrer">golamkabir.com</a>. I'm Golam Kabir, an X (Twitter) growth specialist. Try my <a href="https://golamkabir.com/tools" rel="noopener noreferrer">free X tools</a>.</em></p> growth marketing twitter socialmedia Golam Kabir Sat, 03 Oct 2026 02:43:26 +0000 https://dev.to/iamgolamkabir/how-the-x-algorithm-works-in-2026-1a0h https://dev.to/iamgolamkabir/how-the-x-algorithm-works-in-2026-1a0h <h2> Key takeaways </h2> <ul> <li>The X algorithm people argue about is mostly the For You ranker. Following is a different feed, and X has not published a complete, stable spec for every change to it.</li> <li>The numbers on this page come from the public xAI repository (home-mixer/params/param.rs), checked 26 Sep 2026. They multiply predicted probabilities, not raw like and reply counts.</li> <li>That file has no bookmark weight, no Premium multiplier, and no link-penalty percentage. Do not treat rumors about those as published facts.</li> <li>An X algorithm reset is not a documented per-account switch. A reach drop has several ordinary causes. The open-source file can also lag the live ranker.</li> </ul> <p>“X algorithm” is one of the least precise phrases in this job. People use it for the For You feed, for a reach slump, for a rumor they saw in a reply, and for a feeling that the app is personally annoyed with them.</p> <p>I have advised 500+ accounts over 10+ years. I still will not invent a score X has not published. This page sticks to what X has put in public, plus a clear label on everything that is only a working interpretation. If you want to turn the published weights into a relative illustration, use the <a href="https://golamkabir.com/tools/x-algorithm-simulator" rel="noopener noreferrer">X algorithm simulator</a>. The simulator is not the live ranker. It says so on the page, and it should.</p> <h2> Two feeds, not one machine </h2> <p>X shows a For You feed and a Following feed.</p> <p>For You is the recommendation feed. It can show posts from people you do not follow. That is the system people mean when they say the algorithm. X has described it, across public engineering posts and later open-source code, as a ranker: it tries to predict how likely you are to take certain actions on a post, then orders what you see.</p> <p>Following is the feed of accounts you chose. I am not going to claim it is a pure clock. X has not published a complete, unchanging description of every Following ranking tweak. If a post from someone you follow is missing, that is not automatic proof of a secret penalty. It can be recency, muting, the author’s settings, or a client bug. I do not have a public spec that lets me rank those causes.</p> <p>When someone asks for the Twitter algorithm explained, I start with that split. Advice that mixes the two feeds usually sounds confident and is hard to test.</p> <h2> What “published” means here </h2> <p>There have been at least two public eras.</p> <p>In 2023, Twitter released an earlier recommendation codebase and an engineering write-up. People still quote figures from that era, often from memory, often wrong. I am not repeating those figures on this page. Mixing 2023 blog-post math with 2026 code is how bad charts get made.</p> <p>The weights below are from a later public repository, github.com/xai-org/x-algorithm, in home-mixer/params/param.rs, checked on 26 September 2026. The same note is on the simulator. If you are reading this later, open the file. Do not trust a screenshot of this article as the live configuration.</p> <p>Two limits matter before any number:</p> <p>The public file is a snapshot. The ranker X is running today can be ahead of it, behind it, or partly different. An open-source drop is not a contract that production will match the file forever.</p> <p>The weights multiply the <strong>predicted probability</strong> of an action. They do not multiply the raw number of likes or replies sitting on a post. A reply weight of 5.0 does not mean “one reply is worth ten likes” in your analytics. It means the model’s estimated chance of a reply is scaled by 5.0 inside that scoring function, while the estimated chance of a like is scaled by 0.5. Those are different objects. Treating them as a conversion table is the most common misread I see.</p> <h2> Weights in the public For You file </h2> <p>These are the figures this site already cites from that file. I am not adding new ones.</p> <p>Positive signals, as published:</p> <ul> <li>A reply is 5.0.</li> <li>A like is 0.5. Replies are weighted far more heavily than likes.</li> <li>A repost is 1.0.</li> <li>A quote is 5.0.</li> <li>A follow is 4.0.</li> <li>A copy-link share is 20.0.</li> <li>A reply from someone you follow back adds 15.0 to the reply weight. That is an addition inside the reply term, not a separate promise about your follower count.</li> <li>DM shares are 5.0 in the same public notes.</li> <li>Dwell is 0.05. A video open is 0.07. A photo expand is 0.05. These are small next to a reply.</li> <li>Opening a link has a small positive weight, 0.2, on the predicted probability of that action.</li> <li>Profile clicks are weighted 0.0 in that file.</li> </ul> <p>Negative signals, as published, again on predicted probability rather than raw counts:</p> <ul> <li>A report is -234.</li> <li>Mute author is -58.8.</li> <li>Not interested is -43.2.</li> <li>Block author is -31.2.</li> </ul> <p>The size of the negative weights is the part operators should respect. A handful of strong negative actions can dominate a pile of light positive ones inside this scoring function. That does not mean I can look at your account and count how many mutes you received. X does not show you that. It means the published math punishes predicted negative actions very hard, so posting in a way that attracts blocks, mutes, and reports is a bad trade even when the like count looks fine.</p> <p>Source, again: github.com/xai-org/x-algorithm, home-mixer/params/param.rs, checked 26 Sep 2026.</p> <h2> What the file does not say </h2> <p>This is the section that should slow you down.</p> <p><strong>There is no bookmark weight in the file this site checked.</strong> Bookmarks still matter to you as a creator, because a save means someone wanted the post later. They are in the <a href="https://golamkabir.com/tools/x-engagement-calculator" rel="noopener noreferrer">engagement rate calculator</a> for that reason. They are not, in this public snapshot, a listed For You weight. Do not claim X “counts bookmarks as 50x” or any other multiple. That number is not in the file.</p> <p><strong>There is no Premium multiplier in the file.</strong> Subscribers get product features, including longer posts, that X has marketed separately. I will not attach a ranking percentage to Premium. The simulator leaves Premium qualitative for this reason. If a thread tells you Premium is “a 4x boost” or any other factor, ask for the file and the line. A testimonial is not a weight.</p> <p><strong>There is no link-penalty percentage.</strong> Posts with external links often get less reach in practice, so I still tell clients to test the same idea with the link in the first reply or in the bio. That is a test, not a published bypass. The public code does not say “links lose 70 percent” or any other cut. The link-open weight above is small and positive. A positive weight on opening a link is not the same thing as a penalty for including one. Both can be true in a live system that has other, unpublished filters. I do not know those filters, so I will not describe them.</p> <p><strong>Author replies under your own post are worth doing.</strong> X does not publish a special multiplier for them in the notes this site uses. I still reply, because the conversation is the point and because replies in general are a heavy weight. I do not tell clients that answering their own thread “doubles reach.” I have not seen that claim in the file.</p> <p>If you want to poke the published weights without pretending you are inside X’s production model, the <a href="https://golamkabir.com/tools/x-algorithm-simulator" rel="noopener noreferrer">algorithm simulator</a> applies the reply, like, repost, quote, and media weights to likelihoods you choose, then scales a relative total. Link placement and Premium do not move that number. That limitation is the honest part.</p> <h2> An X algorithm update is not a mood </h2> <p>X algorithm update is what people type after a bad week. Sometimes they are right that something shipped. Sometimes they changed topics, audience, or posting habits in the same week and blamed the feed.</p> <p>Here is a careful way to talk about updates:</p> <p>X can change ranking, filtering, and product rules. Some of those changes are discussed in public posts by X or by the people who run it. Some land in the open-source repository. Some may exist only in production. I cannot see production. You cannot see production. A public commit is evidence of what was published, not a complete diary of the live system.</p> <p>When I review an account after a slump, I look for ordinary explanations before I reach for “update”:</p> <ul> <li>The posts stopped earning replies and started earning only likes, or neither.</li> <li>The topic drifted away from the people who used to respond.</li> <li>External links moved into the main post.</li> <li>The account picked up behavior that attracts mutes and blocks: repetitive pitches, bait, or obvious automation.</li> <li>The comparison window is a viral outlier, so “normal” looks like a crash.</li> </ul> <p>None of those require a secret patch. They are also fixable without a rumor.</p> <p>If you track this for a company, write down the date you changed the content, the date you changed the profile, and the date a public ranking note actually appeared. Otherwise every dip becomes an update story, and you learn nothing.</p> <h2> Is there an X algorithm reset? </h2> <p>X algorithm reset is a phrase creators use. It usually means one of two ideas. They are not the same, and only one of them has a public footing.</p> <p>The first idea is that X replaced or heavily revised the ranker. That kind of change happens. The 2023 release and the later xAI repository are already two different public artifacts. Calling a real ranking overhaul an “update” is fair. Calling it a personal reset of your account is not, unless X says the overhaul wiped per-account state. I have not seen that stated in the file this page cites.</p> <p>The second idea is a folk remedy: your distribution was zeroed, and if you behave a certain way for a set number of days it comes back. I will not publish a day count. X has not, in the material this site relies on, published a reset button, a cooldown length, or a promise that reach returns on a schedule. Anyone selling a “reset protocol” with a timer is ahead of the evidence.</p> <p>What I do when reach falls, instead of hunting a reset:</p> <ul> <li>Read the last twenty posts as a stranger. If you would not follow this account, the ranker is not the first problem.</li> <li>Check whether replies are still happening in other people’s threads. If your own posts are quiet and your replies are also quiet, look at who you are replying to. The method is in <a href="https://golamkabir.com/blog/how-to-find-the-right-people-to-engage-with-on-x" rel="noopener noreferrer">how to find the right people</a>.</li> <li>Compare a post with a link in the body to a post with the link moved. One test. Not a law.</li> <li>Stop automation if you are using it. Bots and pods are a good way to collect the negative actions the published file weights so heavily.</li> <li>Use the simulator to see which actions the public file cares about, then go earn those actions from the right people. A high predicted like on the wrong audience is still the wrong audience.</li> </ul> <p>A slump can last longer than a week for reasons that have nothing to do with a hidden timer. Relevance is slow. That is uncomfortable, and it is still the working explanation I trust more than a reset myth.</p> <h2> What I would not do with this information </h2> <p>I would not build a spreadsheet that converts ten likes into one reply and then calls the result “algorithm score.” The math is not that.</p> <p>I would not delete a month of posts because a thread said the account needs a clean slate. Deleting can remove the proof a new visitor needed. Do it when a post is wrong or off-strategy, not as a ritual.</p> <p>I would not buy engagement to “train” the model. You would be training it on a crowd that does not buy, and you would be courting the negative weights if the pattern looks automated.</p> <p>I would not quote a weight from this article without the date. The file can change. The sentence to use is: as of the 26 Sep 2026 check of param.rs, a reply was 5.0 and a like was 0.5, on predicted probability.</p> <h2> What to do on the account anyway </h2> <p>The practical reading of the public file is almost old-fashioned.</p> <p>Write posts people can answer. A reply is a heavy positive weight. A like is a light one. Ask for the reply only when you actually want the answer. Empty “thoughts?” lines do not create a real conversation, and they annoy the readers who still give you the benefit of the doubt.</p> <p>Stay in a topic long enough that the same people recognize you. The ranker can only predict interest in what you keep demonstrating.</p> <p>Be careful with outrage and with bait that earns reports, mutes, and blocks. The published negatives are enormous next to a like. Even if production differs, I have never seen a commercial account win by collecting blocks from the people it hoped to sell to.</p> <p>Put the profile and the offer in order so a visit from For You can become a follow. Distribution without a clear profile is a tour that ends at the door. That work is covered in <a href="https://golamkabir.com/blog/how-to-grow-on-x-in-2026" rel="noopener noreferrer">how to grow on X in 2026</a> and in the <a href="https://golamkabir.com/resources/x-growth-playbook" rel="noopener noreferrer">X Growth Playbook</a>.</p> <p>If you want a human to look at the account with you, <a href="https://golamkabir.com/services/x-organic-growth" rel="noopener noreferrer">organic X growth</a> and <a href="https://golamkabir.com/services/x-consulting" rel="noopener noreferrer">X consulting</a> are the relevant offers. Neither one includes a private ranking lever. I do not have one. The simulator is the tool. This article is the caution label.</p> <p>SEARCH &amp; AI FAQ</p> <h2> Frequently Asked Questions </h2> <p>Direct practitioner answers to common industry inquiries.</p> <p>How does the X algorithm work in 2026?</p> <p>For You ranks posts by scoring the predicted probability of actions such as replies, likes, reposts, and quotes, using weights from X's published code. A reply is weighted much more heavily than a like in that file. The weights are not a count of how many likes equal one reply. The live system can differ from the public snapshot.</p> <p>Where do the X algorithm weights on this page come from?</p> <p>They come from github.com/xai-org/x-algorithm, file home-mixer/params/param.rs, as checked on 26 Sep 2026. They are not from the 2023 Twitter recommendation release. If the file changes, these figures need to be checked again.</p> <p>Did X reset the algorithm?</p> <p>X has shipped ranking changes, and it has published ranking code. This site has not found a published reset button, a personal cooldown, or a promise that reach returns on a schedule. Treat reset threads as speculation unless X states the mechanism.</p> <p>What is an X algorithm update?</p> <p>An update is a change to how posts are ranked or filtered. Some changes show up in the public repository. Some may ship in production before, or without, a matching public commit. A creator's reach chart is not, by itself, proof of a specific update.</p> <p><em>Originally published at <a href="https://golamkabir.com/blog/how-the-x-algorithm-works-in-2026" rel="noopener noreferrer">golamkabir.com</a>. I'm Golam Kabir, an X (Twitter) growth specialist. Try my <a href="https://golamkabir.com/tools" rel="noopener noreferrer">free X tools</a>.</em></p> growth Golam Kabir Sat, 03 Oct 2026 02:43:14 +0000 https://dev.to/iamgolamkabir/how-to-grow-on-x-in-2026-20o4 https://dev.to/iamgolamkabir/how-to-grow-on-x-in-2026-20o4 <h2> Key takeaways </h2> <ul> <li>How to grow on X in 2026 is the same job as how to grow on Twitter: a clear profile, a few useful posts, and replies in the right conversations.</li> <li>More followers only help if those people can buy, hire, partner, or learn from you. A bought audience does not do that.</li> <li>Fast growth offers usually mean bots, pods, or follow-unfollow. Organic growth is slower because it depends on being useful in public.</li> <li>The practical sequence is profile, audience map, a small posting rhythm, and daily replies. Measure relevance, not a vanity spike.</li> </ul> <p>People still ask me how to grow on Twitter. The product is called X. The job did not change with the name. You want the right people to recognize you, follow you, and eventually trust you enough to talk.</p> <p>I have spent more than ten years on this, and I have advised 500+ accounts. The pattern that holds up is boring. The pattern that gets sold as a shortcut usually breaks the account.</p> <p>This is the sequence I would still use in 2026 if I were starting a founder, a specialist, or a company account from a modest following. It is also the sequence behind the <a href="https://golamkabir.com/resources/x-growth-playbook" rel="noopener noreferrer">X Growth Playbook</a> and the <a href="https://golamkabir.com/services/x-organic-growth" rel="noopener noreferrer">organic X growth</a> work I do with clients. If you want the business version in shorter form, read <a href="https://golamkabir.com/blog/the-organic-x-growth-framework-for-businesses" rel="noopener noreferrer">the organic X growth framework</a>.</p> <h2> What growing on X actually means </h2> <p>How to grow on X is not “get the number up.” A follow is a weak signal. Someone tapped a button. They may never see you again, never buy, and never remember your name.</p> <p>A useful definition is narrower. Growth means more of the right people can find you, understand what you do in one glance, and have a reason to come back. For a company, the right people are buyers, operators, partners, and the peers those buyers already listen to. For a personal account, it is the same list with your name on it.</p> <p>That is why I treat follower count as a lagging clue, not a goal. If the new followers are in the market, the count is fine. If they arrived because a post escaped into a crowd that will never care, the count is a distraction. I will not give you a monthly follower target. I do not have a number that is true for every account, and inventing one would be a sales trick.</p> <p>How to grow on Twitter 2026 searches are full of the opposite promise: a pace, a hack, a bot. Ignore the pace. Keep the work.</p> <h2> Fix the profile before you ask for follows </h2> <p>Most accounts try to grow with the composer open and the profile untouched. A stranger who likes a reply will tap your name. You have a few seconds. If the bio is a list of hobbies, they leave.</p> <p>X profile work is simple, and it is the part people skip because it feels like stationery:</p> <ul> <li>The name field should be a name a human can say, not a keyword stack.</li> <li>The bio should say who you help and what you do. Clever is optional. Clear is not.</li> <li>The banner should still make sense on a phone, next to the avatar.</li> <li>The pinned post should be the best proof you have, not your funniest post from last year.</li> <li>The link should go to one next step, not a menu of twelve.</li> </ul> <p>The <a href="https://golamkabir.com/tools/x-bio-generator" rel="noopener noreferrer">Twitter bio generator</a> is a template, not a writer. Use it to force the 160-character limit, then edit until it sounds like you. If the profile is the actual problem, the <a href="https://golamkabir.com/services/x-account-audit" rel="noopener noreferrer">X profile optimization</a> audit is the longer version: bio, banner, pin, and the link, reviewed the way a first-time visitor sees them. The article <a href="https://golamkabir.com/blog/what-makes-an-x-profile-convert" rel="noopener noreferrer">What Makes an X Profile Convert</a> walks through the same pieces.</p> <p>Do this before you worry about a content calendar. A sharp post that leads to a vague profile wastes the visit.</p> <h2> How to get more followers on X </h2> <p>How to get more followers on X, and how to gain more followers on X, are the same question with two verbs. The organic answer has three parts. None of them is “follow five hundred people and unfollow the ones who do not follow back.”</p> <p><strong>Be findable in conversations that already have an audience.</strong> Your own posts start with almost no distribution when the account is small. Other people’s posts do not. A specific reply under a relevant thread is how a stranger meets you. Generic praise is invisible. A short, concrete addition is not. I wrote the mapping method in <a href="https://golamkabir.com/blog/how-to-find-the-right-people-to-engage-with-on-x" rel="noopener noreferrer">How to Find the Right People to Engage With on X</a>. Use it. Thirty relevant accounts beat a notification on a celebrity.</p> <p><strong>Give them a reason to tap Follow.</strong> The reply earns the visit. The profile earns the follow. If either is weak, you will feel busy and stay small.</p> <p><strong>Post so the new follower has something to read.</strong> Two or three posts a week that teach something from your actual work will beat a daily queue of quotes. <a href="https://golamkabir.com/blog/why-posting-more-isnt-the-answer-on-x" rel="noopener noreferrer">Why posting more is not the answer</a> is the longer argument. The short version: volume without a point of view trains nobody.</p> <p>Grow Twitter followers organically by repeating that loop. Reply, get the visit, earn the follow, give the follower a timeline worth staying for. Then do it again tomorrow. That is the whole method. The rest is taste and consistency.</p> <h2> A week that does not depend on inspiration </h2> <p>I do not believe in a perfect posting time as a growth strategy. Timing can help a post that is already worth reading. It cannot rescue a post that is not. If you want a planning aid, the <a href="https://golamkabir.com/tools/x-growth-calculator" rel="noopener noreferrer">growth estimator</a> on this site is a calculator, not a forecast of your account.</p> <p>A week I can defend looks like this:</p> <ol> <li>One longer piece. A thread or a tight post that explains a real problem, a method, or a mistake. If the draft is long, the <a href="https://golamkabir.com/tools/x-thread-splitter" rel="noopener noreferrer">Twitter thread splitter</a> keeps each post inside 280 characters so you are editing ideas, not counting on your thumbs.</li> <li>One or two shorter observations from work you actually did. Not a recap of someone else’s thread. Your sentence.</li> <li>A daily block of replies. Twenty to thirty minutes is enough if the list is curated. More time is fine. Random scrolling is not the same thing.</li> <li>A pass on the profile once a month, not every morning. Change it when the offer changes, not when you are bored.</li> </ol> <p>Run the first line of the long piece through the <a href="https://golamkabir.com/tools/x-hook-scorer" rel="noopener noreferrer">hook scorer</a> if you tend to bury the point. A hook is a promise you have to keep in the next lines. It is not a trick.</p> <p>If you cannot write that week yourself, that is a capacity problem, not a strategy problem. <a href="https://golamkabir.com/services/x-ghostwriting" rel="noopener noreferrer">X ghostwriting</a> exists for drafts in your voice, with your approval. <a href="https://golamkabir.com/services/x-content-strategy" rel="noopener noreferrer">Content strategy</a> exists for the pillars before the drafts. Do not hire either one to imitate a meme account.</p> <h2> How to grow on X fast </h2> <p>How to grow on X fast, and how to grow an X account fast, are popular searches because waiting feels like failure. I am going to be plain about what “fast” usually means.</p> <p>Bought followers are fast. They do not read, they do not buy, and they teach the account nothing. Follow-unfollow is fast until the account looks desperate and the real people notice. Engagement pods are fast inside the pod and quiet everywhere else. I do not use any of those, and I will not teach them.</p> <p>There is a legitimate way to move faster than “post and hope.” It is not a hack. It is leverage:</p> <ul> <li>Reply where the readers already are, instead of waiting for your own posts to be discovered.</li> <li>Make one strong public asset (a pinned breakdown, a playbook, a clear offer) so each new visitor has a next step.</li> <li>Stop splitting attention across networks you do not intend to run. This site is about X only for that reason.</li> <li>Cut topics that attract everyone and convert no one.</li> </ul> <p>Even then, organic growth is measured in weeks and months of repetition, not in a weekend. If someone promises a follower count by Friday, assume the followers are not customers. I do not guarantee a number. Anyone who does is selling a different product than the one on this site.</p> <p>The accounts that feel “sudden” from the outside usually had a quiet period of replies first. You are seeing the week it became visible, not the month that made it possible.</p> <h2> What to pay attention to </h2> <p>Open analytics, then ignore most of it. I look at a short list:</p> <ul> <li>Are profile visits coming from replies, or only from your own posts? If replies produce no visits, the replies are too soft or they are in the wrong threads.</li> <li>When someone visits, do they follow? If they do not, the bio and the pinned post are the first place I look.</li> <li>On the posts that do get distribution, are people replying, or only liking? A like is easy. A reply means the post gave them something to add. The published For You weights treat those actions very differently. Read <a href="https://golamkabir.com/blog/how-the-x-algorithm-works-in-2026" rel="noopener noreferrer">how the X algorithm works</a> before you treat every interaction as equal, and use the <a href="https://golamkabir.com/tools/x-engagement-calculator" rel="noopener noreferrer">engagement rate calculator</a> if you want the arithmetic.</li> <li>Are the new followers in the market? Spot-check bios. A random sample tells you more than a rounded total.</li> </ul> <p>The <a href="https://golamkabir.com/tools/x-account-health-checker" rel="noopener noreferrer">account health check</a> is a habit checklist. It does not log into X, and it does not tell you whether the account is limited. Use it to notice link placement, reply habits, and automation. Then go look at the account yourself.</p> <p>If the numbers are flat and the work is real, the usual causes are a vague offer, replies in the wrong neighborhood, or posts that could have been written by anyone in the category. Fix one of those before you add a fourth post per day.</p> <h2> Where the playbook fits </h2> <p>The <a href="https://golamkabir.com/resources/x-growth-playbook" rel="noopener noreferrer">X Growth Playbook</a> is the free guide: profile, positioning, research, and a daily routine. This article is the 2026 field note around it. They overlap on purpose. The playbook is the checklist. This page is the reasoning, including the parts I will not dress up.</p> <p>If you want the work done with you, the <a href="https://golamkabir.com/services/x-organic-growth" rel="noopener noreferrer">X growth service</a> is research, positioning, content, and human replies. It is not a bot. It is not a follower package. <a href="https://golamkabir.com/services/audience-research" rel="noopener noreferrer">Audience research</a> is the piece to buy first if you do not yet know who should see the account. <a href="https://golamkabir.com/services/x-engagement" rel="noopener noreferrer">Engagement strategy</a> is the piece to buy if the posts exist and the replies are still going to the wrong people.</p> <h2> A simple order, if you are starting this week </h2> <p>Do not try to install the whole system on Monday.</p> <p>Day one, rewrite the bio until a stranger knows who you help. Pin one post you would still defend.</p> <p>Day two, write a private list of about twenty-five accounts: peers, a few larger specialists, and actual buyers. Not celebrities.</p> <p>Day three, reply twice in that list with something specific. No “great post.”</p> <p>Day four, publish one post from your own work. One idea. Stop.</p> <p>Then repeat the replies every weekday and the original post a few times a week. Review the profile in a month, not tomorrow.</p> <p>That is how to grow on X in 2026. It is also how to grow on Twitter, because your buyers are still using both words for the same place. The accounts that get this right look obvious in hindsight. They were just specific, in public, for longer than the shortcut sellers were willing to be.</p> <p>SEARCH &amp; AI FAQ</p> <h2> Frequently Asked Questions </h2> <p>Direct practitioner answers to common industry inquiries.</p> <p>How do you grow on X in 2026?</p> <p>Make the profile obvious, post a small number of specific ideas, and spend more time replying in conversations your buyers already read. That is how to grow on X and how to grow on Twitter in 2026. Skip bots, follower packages, and engagement pods.</p> <p>How do you get more followers on X without buying them?</p> <p>Earn them. Reply with something a stranger can use, then give that stranger a profile that explains who you help. The follow happens after the reply, not before a pitch.</p> <p>How fast can you grow Twitter followers organically?</p> <p>There is no honest timetable. The first weeks are often quiet because a new account has little distribution of its own. Speed comes from showing up in existing conversations, not from posting more into an empty following.</p> <p>Does posting more grow an X account?</p> <p>Only after the profile and the reply habit are in place. Extra posts without an audience or a point of view mostly create a longer timeline nobody asked to read.</p> <p><em>Originally published at <a href="https://golamkabir.com/blog/how-to-grow-on-x-in-2026" rel="noopener noreferrer">golamkabir.com</a>. I'm Golam Kabir, an X (Twitter) growth specialist. Try my <a href="https://golamkabir.com/tools" rel="noopener noreferrer">free X tools</a>.</em></p> growth