How the X Algorithm Works in 2026 (And How Founders Can Win)

The X algorithm is open-source. You can read the exact code on GitHub. You can see the weight multipliers for every engagement type. You can measure how much each signal drives reach. Most founders know this. Few act on it.

Study the algorithm first, then build strategy around what the code actually rewards. Create content and hope the algorithm picks it up and you lose. This is the difference between engineering distribution and renting attention from luck.

The 3-Stage Pipeline: How Posts Get Ranked

X processes 500 million tweets every day. The algorithm selects about 1,500 candidates for your feed. Those candidates get ranked using a neural network trained on billions of engagement signals. The final ranking applies heuristic rules to filter spam, enforces diversity, and surfaces fresh content from your network.

This pipeline makes 5 billion ranking decisions daily. Half come from people you follow (Real Graph). Half come from people you don't follow but post about topics in your interests (SimClusters). The balance shifts based on your behavior, account age, and Premium status.

Your job is not to game every signal. Understand which signals matter most and build content that consistently triggers them. The algorithm is transparent. Act like it.

The Exact Engagement Weights (From Open-Source Code)

Every engagement type has a weight multiplier. Here's what each action is worth:

The highest-value actions:

Reply engaged by author (conversation): +75 weight. This is 150x more valuable than a like. Someone replies to your tweet, you reply back. That creates a conversation signal that compounds your reach. This is the single most powerful lever on X.

Reply (any reply): +13.5 weight. That's 27x a like. A reply shows the algorithm your content triggered a response. Replies move posts into the Conversation section. They get more visibility than likes because they signal engagement.

Profile click plus engagement: +12.0 weight. Someone clicks your profile and engages with something there. The algorithm records a strong intent signal.

Conversation click plus engagement: +11.0 weight. Your tweet sparked enough interest that people click into the thread. They're reading deeper, not just reacting.

Dwell time (2+ minutes): +10.0 weight. Someone reads your tweet long enough to read it twice or think about it. The algorithm watches how long a tweet stays on someone's screen.

Bookmark: +10.0 weight. Bookmarks are save signals. "I want to read this later" or "I want to reference this." More valuable than a like because it shows high intent.

Retweet: approximately 20x a like. A retweet tells the algorithm your content matters enough that someone wants to broadcast it to their network. This extends your reach beyond your followers.

Like: +0.5 weight. This is your baseline. Likes are the least valuable action. They're easy and low-intent. A like takes one-tenth of a second. Don't optimize for likes.

Negative signals (the reach killers):

Block, mute, or report: approximately -74x weight. If someone blocks you, mutes you, or reports your content, the algorithm cuts your reach by about 74x. More critically, that person's account is removed from your audience entirely. One block doesn't kill a post. Multiple blocks on the same content type tells the algorithm to stop showing that content.

The Grok Rebuild: What Changed in January 2026

In January 2026, X's engineering team rebuilt the entire ranking system in Rust. The old architecture was Scala with technical debt from a decade of patches. The new system is modular, fast, and uses a transformer-based neural network called Grok.

The architecture includes Home Mixer (decides in-network vs. out-of-network), Thunder (in-memory storage), Phoenix (the Grok ranking model), and the Candidate Pipeline (sources 1,500 candidates).

Grok analyzes sentiment on every post. Positive, constructive content gets wider distribution. Negative, inflammatory content gets suppressed even if it gets early engagement. Different from the old system, which was pure engagement-based. The new system has editorial judgment.

Grok also enables promptable feeds. You give natural language commands to filter your feed. "Show me startup founders" or "Hide cryptocurrency content." This changes what gets selected into your personal feed from the broader algorithm.

X updates the code on GitHub every 4 weeks. The repository got 1,600 GitHub stars in the first 6 hours. The transparency is real. Read the code.

Grok processes 100 million videos daily. It analyzes visual sentiment and extracts meaning from content. Video with strong sentiment scoring gets wider distribution.

TweepCred: The Hidden Reputation Score Controlling Your Reach

TweepCred is a hidden score between 0 and 100. X doesn't publish it, but the code shows the factors that build it.

The critical threshold is 65. Below 65, you're ineligible for broad distribution. Only your direct followers see your posts. Only 3 of your tweets are eligible for the For You feed. Above 65, your account is healthy. Above 75, you get distribution boosts.

Your TweepCred builds from account age, follower-to-following ratio, engagement quality, and engagement with your network. A Premium account gets a +4 to +16 boost to stay above 65.

To improve TweepCred, get followers intentionally. Quality followers (verified accounts, engaged users, domain experts) help more than quantity. Write content that gets replies. Reply to other people's posts. Engage with your community.

Don't buy followers. Don't engage with spam. Don't like random content to appear engaged. The algorithm sees these tactics and punishes them.

SimClusters and Real Graph: How X Knows Your Audience

X has 145,000 topic clusters called SimClusters. "Startup founders" is a cluster. "Artificial intelligence" is a cluster. "Personal finance" is a cluster. Your account belongs to multiple clusters based on what you post and who you engage with.

The algorithm uses SimClusters to route your content to people interested in your topic, even if they don't follow you. Post in the "startup founders" cluster and the algorithm routes it to other accounts in that cluster on their For You feed.

Real Graph is the in-network recommendation system. It looks at who you follow and what you interact with, then shows content from your extended network.

Topical consistency is key. Post about startups 80% of the time, random content 20% of the time. The algorithm assigns you primarily to the startup cluster. Your reach within that cluster is 3x higher than if you post equally across topics.

Jump from startups to cryptocurrency to fitness and the algorithm fragments your cluster assignment. You're weak in multiple clusters instead of strong in one.

Own a single SimCluster. Pick the one aligned with your business. Post about it 80% of the time. Reply to others in that cluster. Build relationships with strong voices. Become the default recommendation when someone new enters.

Out-of-network distribution comes from being strong in your SimCluster. The algorithm shows your content to new people in that cluster because you're a strong voice.

Premium vs Free: The 10x Reach Gap

This is the number most founders miss. Free accounts get less reach. Premium accounts get more. The gap widened dramatically in 2026.

Free account median reach: less than 100 impressions per post. If you post to 1,000 followers, you're not guaranteed 1,000 impressions.

Premium reach: 600+ impressions per post. That's 10x. Precisely: Premium accounts get 4x in-network boost and 2x out-of-network boost.

Premium+ reach: approximately 15x free accounts. Premium+ adds creator fund eligibility and algorithmic boosts.

Since March 2026, the algorithm applies an aggressive penalty to free accounts posting external links. A free account posting a link gets zero median engagement. The algorithm doesn't surface those posts. This is enforcement around link penalties.

For a founder, Premium is not optional if you're building an audience on X. The ROI is straightforward. Post twice a day, that's 60 posts per month. At 600 impressions per Premium post versus 80 impressions per free post, you get 31,200 extra impressions per month. Premium costs $168 per year. That's about $14 per month.

Your customer acquisition cost is probably $50 or more. If even 0.05% of those extra impressions convert to something valuable, Premium pays for itself in the first month.

The Link Penalty: How Founders Work Around It

External links get a 30-50% reach reduction in the code. In real-world testing in Q1 2026, the penalty is more severe. Posts with links see up to a 94% reach reduction compared to posts without.

This isn't a bug. It's designed penalty. X wants engagement to stay on X. Everyone posts links to external blogs and users leave X. The algorithm punishes this.

For founders building thought leadership, this creates a problem. You want to link to your blog and products. The algorithm doesn't want you to.

Three workarounds exist.

The reply-link strategy: Post your insight as text-only. Get engagement on the text post. Once the post has replies and momentum, reply to your own post with the link. The algorithm gave the text post reach. Adding a link in a reply doesn't trigger the penalty the same way.

The article feature: X has a native article feature. Write long-form content directly on X without external links. The algorithm treats native articles favorably because they keep users on X. Use the article feature instead of linking to your blog.

The strategic link post: Accept that link posts get lower reach. Treat them as retention, not acquisition. Your followers see them. New people don't. Don't expect algorithmic reach. Use direct outreach, email, or ads to drive traffic. Save algorithmic reach for text-only posts that build authority.

Content Format Rankings

X is the only major platform where text outperforms video. This matters for founder strategy.

Text-only: 30% more engagement than video, 37% more than images. Text allows nuance, clarity, precision. Text formats well on mobile. Text compounds through replies and conversations.

Video: 5.4% more than images, 18% more than links. Video feels like it should perform better, but it doesn't. Most videos are consumed passively. They don't trigger reply and conversation signals that drive reach.

Images: Lower engagement than text and video. Work for context (product, chart), not for driving high-value signals.

Links: Lowest performance due to the penalty.

Character count matters. Tweets between 71-100 characters get 17% higher engagement. Tight, punchy content. Headlines, not essays.

Threads (4-8 connected tweets) perform better than single long posts for certain content. They break complex ideas into pieces. Each tweet acts as a separate ranking candidate, giving multiple shots at different audiences.

Long-form single posts (using X's article feature or a single very long tweet) get 40-60% more impressions than threads. Longer content formats, when native to X, get algorithmic preference.

The ideal format for founders: short, punchy text posts, 71-100 characters, designed to trigger replies. Use the article feature or a very long single post for longer content.

30 Algorithm-Aware Tactical Plays for Founders

Specific, executable plays based on algorithm weights and signals.

Reply-optimized plays:

1. End every post with a question. Questions trigger replies. Replies trigger the 150x conversation signal. "What's the biggest mistake you see founders make?" outperforms "Founders make mistakes" every time.
2. Reference something from the previous hour on your topic. Reply within 60 minutes. The algorithm is freshness-biased. Recent replies on recent posts get high visibility. Get in early on conversations in your SimCluster.
3. Ask why instead of what. "Why do founders hate fundraising?" gets 3x replies of "What should founders know?" Why questions invite opinion. What questions invite facts.
4. Reply to other people's posts in your cluster before posting your own. The algorithm boosts accounts that participate in conversation. Reply to 5 posts, your next original post gets better distribution.
5. Pin a high-engagement post and make it your first conversation hook. New visitors see your best post. Make that post ask a question or invite a strong take. New followers reply if the first thing they see is controversial.
6. Create a "hot takes" post that's intentionally debatable. "Most founders should not raise venture capital." This triggers strong replies. Disagreement counts as engagement. The algorithm doesn't differentiate. Controversy compounds reach.
7. Respond to every reply on your posts for 4 hours after posting. Early replies compound. The algorithm weights fresh engagement higher. Fast replies trigger the conversation chain signal multiple times.

Bookmark and dwell-time plays:

8. Design posts for re-reading. Use formatting, line breaks, structure. Make posts that require a second read. Dwell time is how long someone spends reading. Clear formatting invites longer reads.
9. Post lists (3-item, 5-item, 7-item). Lists trigger bookmarks. "3 things founders get wrong about product-market fit" formats well and invites bookmarks.
10. Include a counter-intuitive statement in the first line. "Passion is overrated." The reader re-reads to understand the contradiction. Dwell time increases.
11. Build a small series (same topic, same time daily). "Founder failure breakdown" every Tuesday. Followers bookmark to collect. Consistency plus topical coherence equals bookmark magnet.
12. Post insights that answer common questions in your field. "Here's why series A funding destroys startups" answers a question people search for. People bookmark to save the answer.

Engagement and profile-click plays:

13. Build a 2-sentence bio that describes one specific outcome. "I help technical founders build $1M ARR businesses." Profile visitors click through when they see specificity.
14. Link your bio to a high-converting resource. Landing page, free guide, course signup. Profile clicks matter only if the follow-through is good.
15. Create a polarizing profile description. "Founder who chose bootstrapping over venture." Specific enough to attract your audience. The algorithm loves relevance.
16. Post case studies with specific numbers. "$0 to $50K ARR in 90 days." Specific numbers get clicked and bookmarked.

Topical consistency and SimCluster plays:

17. Do a quarterly SimCluster audit. Analyze your posts from the last 90 days. What percentage are in your core topic? Below 80% means you're diluting your cluster assignment.
18. Create a 10-tweet "master thread" once per quarter that defines your area of expertise. "Everything I know about founding." This becomes a reference post that gets shared repeatedly.
19. Create content that bridges adjacent clusters naturally. "How startup marketing is just psychology applied to business." This connects the startup cluster and the psychology cluster. You own both.
20. Reply exclusively in your cluster for 30 days. Only engage with people posting about your topic. The algorithm adjusts your cluster assignment. After 30 days, your out-of-network reach increases.
21. Identify the 5 strongest voices in your SimCluster. Reply to every post they publish. The algorithm notes that you're in conversation with cluster leaders. You inherit some of their strength.

Frequency and timing plays:

22. Post exactly 2 times per day, 8 hours apart. More frequent posting doesn't increase reach proportionally. The algorithm sees volume as noise. 8 hours apart catches different time zones.
23. Post at 7am, 2pm, or 9pm EST. These times hit peak activity windows. The algorithm gives early engagement more weight.
24. Never post 3 times in a 4-hour window. The algorithm treats this as spam. TweepCred takes a hit. Consistency matters more than volume.
25. Post your best content on Tuesday or Wednesday. Weekday engagement is 3x weekend. Wednesday is peak for professional content. Save strongest ideas for midweek.

Negative-signal avoidance plays:

26. Never reply to content you disagree with strongly. Disagreement replies attract people who don't like you. They're more likely to mute or block. Avoid low-quality discourse.
27. Don't engage with accounts below your follower count by 10x. Small accounts engaging with huge accounts look like bot behavior. Keep engagement concentrated on similar-sized or larger accounts.
28. Keep ratio below 5. Ratio is (likes + retweets) / replies. Disproportionate likes and low replies means you're broadcasting. The algorithm downgrades reach.
29. Mute words related to your industry's controversies. If your industry is getting roasted, don't participate. Stay focused on your specific angle.

TweepCred building plays:

30. Follow 10 new high-quality accounts in your space each month. Following people in your SimCluster signals you're involved. Your engagement improves TweepCred.
31. Write 1 substantive reply per day to someone with 10x+ your followers. Replies to large accounts get visibility. High-quality replies boost TweepCred because engagement with bigger voices signals credibility.
32. Get 10 new followers per week from your cluster. Quality followers boost TweepCred more than quantity. Engage deliberately with people you want to follow. "I write about X. You'd find value if you follow."

The Conversation Flywheel: Engineering Reply Chains

The conversation signal is +75 weight (150x a like). This is the nuclear button on X. Most founders never trigger it because they post and disappear.

The conversation signal fires when someone replies to your post and you reply back. The second reply triggers the +75 multiplier. It signals a two-way conversation, not a monologue.

To engineer conversation flywheels:

Step 1: Ask a specific question in your post. "What's the biggest mistake you made as a first-time founder?" This is specific and invites varied answers. Post at peak time. Respond to every reply within the first hour. Reply to all of them.

Step 2: Reply with a follow-up question or thoughtful comment. Don't say "Great point." Engage intellectually. "That's real. How did you overcome it?" Your reply invites another reply.

Step 3: Let the conversation run for 4-6 hours. Let other people add to it. Once engagement slows, reply one more time to a new voice. This restarts the engagement clock.

Step 4: Repeat with a different question tomorrow. Same time, same process. The algorithm sees you as a conversation starter. Your posts get better distribution because they drive engagement.

This works because it's authentic. You're actually engaging. The algorithm rewards this because engagement is what X sells to advertisers.

The Founder's Algorithm-Optimized Content Calendar

A weekly schedule designed around algorithm signals.

Monday: Core expertise post. A substantive insight in your area of expertise. Timeless advice. "The main reason early-stage startups fail is misalignment between co-founders." Posts at 7am EST. This is your SEO post, your searchable content.

Tuesday: Controversial take. Something that gets disagreement and discussion. "Venture capital is overrated." Posts at 2pm EST. Controversy drives replies. Respond to every disagreement actively.

Wednesday: Case study or data. "$0 to $100K ARR in 6 months. Here's how." Specific numbers and results. Gets bookmarked and shared. Posts at 7am EST.

Thursday: Question-based engagement thread. "What's the biggest lesson you learned about fundraising?" Posts at 2pm EST. You respond to every reply. This is your conversation flywheel day.

Friday: Hot opinion or pattern you're seeing. "Every founder I talk to is worried about recession. None of them are actually preparing for it." Posts at 9pm EST (end of work week). Gets engagement from people reflecting.

Saturday: Weekend reader. Low-pressure content. "The best book I read this month is X." Posts at 7am EST. Weekend reach is lower, optimize for engaged followers.

Sunday: Minimal posting. Maybe one reply to a trending conversation, but no original posts. The algorithm sees Sunday posting as less serious. Save your budget for weekdays.

This schedule uses each day's strongest format for that audience. Monday insights for people starting the week. Wednesday data for proof. Thursday conversations for engaged people. Friday opinions for people decompressing.

Negative Signals and Recovery

The algorithm penalizes several behaviors. Understanding them helps you avoid them and recover if you trigger them.

Blocks and mutes: 20 people block you in a week means the algorithm sees something wrong. It's not an immediate ban. Audit the 20 tweets right before the blocks. What topic triggered the negative reaction. Avoid similar content. It takes about 2 weeks of clean content for the algorithm to reset.

Content penalties: A single tweet reported multiple times gets reviewed by X's moderation team. If it violates rules, it gets labeled or removed. If it doesn't violate rules but got reported anyway, the algorithm still suppresses reach while it investigates. Don't post anything vaguely reportable. Stay focused and professional.

Unfollows: A few unfollows per week are normal. 50+ unfollows after a tweet signals something. Something didn't match your audience's expectations. Your audience is self-selecting. Some people are wrong for you. Let them go.

Off-topic excursions: Post about a completely different topic and the algorithm penalizes reach by about 3x. Your followers follow you for your expertise. Stay on topic. If you want to post off-topic, use a separate account.

The penalty recovery timeline:

Days 1-3: Post normally on-topic. Respond to replies. The algorithm sees continuity and starts reversing the penalty.

Days 4-7: Increase reply activity. Get engagement from quality accounts. This signals health. Penalties continue easing.

Days 8-14: You're essentially recovered. Your reach is back to normal.

Don't overcompensate. Don't post 10x more content trying to recover. Act normal. The algorithm recovers accounts quickly if the account shows good-faith behavior.

How to Dominate Your SimCluster

You own a SimCluster by becoming the most-shared, most-replied-to, most-bookmarked voice in that cluster.

Step 1: Identify your exact cluster. X has 145,000 clusters. You probably belong to 3-5. "Startup founders," "technical founders," "marketing," maybe "venture capital." What's the smallest, most specific cluster you can own? Don't go for "technology." Go for "early-stage SaaS founders." Specificity equals less competition.

Step 2: Commit to 80%+ of your content being in that cluster for 90 days. This is the reset period. Your cluster assignment shifts based on volume of topical content. After 90 days, your out-of-network reach increases within that cluster.

Step 3: Build relationships with the 5-10 strongest voices in your cluster. Reply to their posts regularly. Reference their ideas. Get them to reply to you. The algorithm sees these relationships and starts cross-recommending.

Step 4: Create a signature content format in your cluster. "Founder failure breakdown," "Marketing lever of the week," "Technical founder problem of the day." A recognizable format unique to you. The algorithm and your followers start expecting it.

Step 5: Own one specific angle no one else covers. If 50 people post about fundraising, find the uncovered angle. "How to fundraise as a solo founder," "Fundraising for bootstrapped companies." Own the gap. The algorithm rewards specificity.

Step 6: Create cross-cluster bridges. You're in the "startup founders" cluster but startups need marketing. Create content that bridges both: "How to do founder-led marketing that scales." You get distribution in both clusters.

Algorithm Myths Founders Still Believe

Myth 1: Posting frequently increases reach. False. High-volume posting looks like noise. 10 times per day doesn't get 10x reach. It gets maybe 2x and your TweepCred takes a hit. 2x per day is optimal. Quality over frequency.

Myth 2: Retweets are better than replies. Retweets are approximately 20x a like. Replies are approximately 27x. Replies are better. Replies trigger the conversation signal which is more valuable.

Myth 3: Hashtags help reach. Hashtags create a 40% penalty if you use more than 2. One or two relevant hashtags give a +21% boost. Most founders use 5-10 and don't understand the penalty. Skip hashtags.

Myth 4: Timing doesn't matter. Timing matters a lot. Posts at peak times (7am, 2pm, 9pm EST) get 3x engagement. Peak time posts compound faster.

Myth 5: Longer threads perform better. Long-form single posts (native articles) perform better for most content. A single long post gets 40-60% more impressions than a 4-tweet thread. Use threads for storytelling. Use long posts for teaching.

Myth 6: Premium is optional. Premium is not optional if you're serious about reach. Free accounts get 10x less reach. Premium pays for itself in the first month.

Myth 7: Likes matter. Likes are vanity. A post with 100 likes and 10 replies gets less reach than a post with 20 likes and 50 replies. Stop optimizing for likes.

Myth 8: Following more people helps reach. False. The algorithm is not affected by your follow count. It's affected by your follower count, engagement quality, and TweepCred. Follow people you want to learn from, not people you want to follow you.

Premium ROI for Founders

Premium costs $168 per year ($14 per month). Is it worth it?

Calculate your current reach:

2 posts per day, 60 posts per month. Average 80 impressions per post as a free account. Total: 4,800 impressions per month.

Premium increases reach 10x. 4,800 times 10 equals 48,000 impressions per month.

Difference: 43,200 extra impressions per month.

What does 43,200 impressions convert to:

0.05% convert to a click: 21 clicks per month. 0.1% convert: 43 clicks per month. 1% convert to engagement: 432 engagements per month.

Your customer acquisition cost is probably $50 plus. If even 1 in 2,000 impressions leads to a customer, Premium paid for itself. The actual conversion rate is probably 10-100x higher.

The real math:

Premium is not optional for any founder serious about distribution. The ROI is infinite. You're paying $14 per month. You'll get at least $50 in value.

Pay for Premium.

Premium+ ROI

Premium+ costs $168 per month. This is a different calculation.

Premium+ gives you approximately 15x the reach of a free account (versus 10x for Premium). The extra 5x is significant but costs $168 per month instead of $14.

Premium+ makes sense if you're already getting consistent traffic and conversions from X, using X as a primary customer acquisition channel, actively monetizing, or have an audience of 50K plus.

For earlier stages, Premium is enough. Don't spend on Premium+ until you've validated that X is driving revenue.

Connect Everything to Business Outcomes

The algorithm is powerful. But power means nothing if it doesn't connect to your business.

Every play should drive one of these outcomes:

Traffic: Get people to click to your site. Focus on profile clicks and conversation engagement. Build a clear link in your bio. Reply to every comment with a thoughtful response.

Email list: Capture emails. Use profile clicks and link-outs to landing pages. Use reply-link strategy: text post gets engagement, reply with link to landing page. Premium accounts get better link visibility.

Sales conversations: Inbound DMs and meetings. Establish authority in your SimCluster. Post about your specific problem. Reply to everyone who engages. Create urgency. Make DMing you feel natural.

Content amplification: Reach on your blog. Own your SimCluster so completely that when you post about your topic, you get massive engagement. Engagement drives impressions. Impressions drive traffic. Start with SimCluster dominance.

Thought leadership: Authority. Consistent, high-quality replies. Responses to others' posts more than original content. Become the person people ask for advice. Doesn't drive direct revenue, but shortens sales cycles and raises deal sizes.

Pick one goal. Optimize for it. Everything else is distraction.

What Founders Actually Get Wrong

You've read the algorithm. You know the weights. You know the signals. You know the plays. Here's what founders get wrong:

Most founders understand the algorithm intellectually but refuse to act on it. They read that replies are 27x more valuable than likes. Then they post polished content and expect people to engage. They don't ask questions. They broadcast instead of having conversations.

The algorithm is transparent. The code is public. Founders who win study the code. Founders who lose know about the code but don't change their behavior.

Second, most founders think about reach as an output. It's an input. Your reach is determined entirely by how well you align with algorithm signals. You don't get reach because you're smart. You get reach because you triggered the +27 reply signal or the +10 bookmark signal. The algorithm doesn't care how good your ideas are. It cares how much they're engaged with.

Third, most founders are impatient with SimCluster dominance. You don't own a cluster overnight. You own it through 90 days of 80%+ topical consistency. Most founders post about their topic for 2 weeks, get impatient, and then post random content. They restart the 90-day clock. Then they're confused why they never get out-of-network reach.

Finally, most founders don't understand that the conversation flywheel is engineering, not luck. You're not hoping people reply. You're designing posts that mathematically trigger replies. You're responding in a way that mathematically triggers more replies. It's science.

Read the code. Study the weights. Build strategy. Execute. Track results. Act like it.

What Comes Next

You've got the playbook. You understand the pipeline. You know the signals. You know the weights. You know how to trigger the 150x conversation signal. You know how to own your SimCluster. You know that Premium is required.

The next step is execution. Pick three plays from the 30 listed. Commit to them for 30 days. Measure results. Track engagement, reach, and traffic impact. Then scale what works.

For more on execution, read /blog/x-growth-playbook-for-founders/. For building in public on X, read /blog/build-in-public-on-x-founders/. For multi-platform authority building, read /blog/founder-thought-leadership/.

The algorithm is open-source. Your strategy should be too.