guides · cadence

How often to post on X

Two to three posts a day and fifteen to twenty five replies. That is the answer, and the rest of this page is why - why volume works, why variety is not optional, why replies scale on a different curve from posts, and how to run all of it without quitting in three weeks.

Why volume works at all

Nobody, including people with large accounts, can reliably predict which post will do well. The post you laboured over dies quietly; the throwaway line you almost deleted takes off. This is not a failure of skill. Whether a post travels depends on who happened to be online, what else was happening, what the first twenty readers did with it, and a ranking system you cannot see.

If outcomes are that noisy, the rational response is more attempts. Say roughly one post in fifteen escapes your follower bubble and finds an out-of-network audience. At one post a day that is two breakouts a month. At three a day it is six. You did not get better at writing; you bought three times as many lottery tickets in a lottery where the tickets are free.

Volume has a second effect. Each post is a data point about what you are about and who responds to you. A quiet account gives the ranking system almost nothing to work with. An active one gets its audience targeting sharpened continuously. This is why accounts often feel like they "click" after a few weeks of consistency: the system finally knows who to show you to.

Variety is the part people skip

Volume without variety backfires, and it backfires through your own followers.

Every post is shown first to some slice of the people who follow you, and what they do in the first stretch decides whether it goes further. If you post the same format five times a day, your followers learn the shape of your posts and start scrolling past on sight. Early engagement drops. Which means fewer of your posts break out - including the good ones. You have trained your own audience to ignore you, and the algorithm reads that as a verdict on your content.

So rotate. A rough weekly deck worth dealing from:

• A lesson - something you learned the hard way, stated plainly.
• A number - your data, an experiment result, a before and after.
• An opinion - a position in your niche that some people would argue with.
• A build update - what you shipped and what it actually changed.
• A question - a real one, with enough context to be answerable.
• A resource - a compact how-to or list people will save.
• An observation - something you noticed that nobody has named yet.

Two or three different shapes a day keeps you unpredictable in the good way. What to post on X goes deeper on each format.

Do not run the winner into the ground

When a post does well, the instinct is to make four more exactly like it. It nearly always disappoints, and the reason is worth understanding.

Part of why that post worked is that it was new to your audience. A format has novelty value that depletes with each repetition. The second version lands softer, the third softer still, and by the fifth you have taught people that this is just what you do now. Meanwhile you have spent a week of posting slots on a decaying asset instead of finding the next thing.

Extract the principle, not the template. If the post worked because it was uncomfortably specific, be uncomfortably specific about something else. If it worked because you disagreed with the consensus, disagree with a different consensus. Repeat the property, vary the form.

Replies scale on a different curve

Posts and replies are both content, and they behave nothing alike.

Your post goes to your followers first. Your ceiling is therefore set by how many followers you have, which is exactly the thing you are trying to increase. At a small account this is a closed loop: you need reach to get followers, and you need followers to get reach.

A reply breaks the loop, because it lands in someone else's audience. Its ceiling is set by their reach, not yours. This means replies are one of the few things on X where doing more of it straightforwardly gets you more, with no follower prerequisite. Twenty replies is twenty appearances in front of people who have never seen you.

The practical consequence for cadence: below a few thousand followers, put more of your daily budget into replies than posts. Two or three posts and twenty replies, not ten posts and two replies. Posts convert people once they arrive; replies are what makes them arrive. The mechanism is laid out in how to get more impressions on X and the reply playbook.

One caveat on reply volume: quality has a floor. Twenty real replies compound; a hundred empty ones read as bot behaviour and get treated as such. And never automate the send - X blocked API-based reply automation in February 2026 and the account at risk is the one you are growing.

Timing matters less than you think, until it doesn't

Posting time is a multiplier on a base number. If your base is small, tuning the multiplier is a rounding error, which is why obsessing over the perfect hour before you have volume is wasted effort. Get to a consistent daily cadence first. Once you are posting regularly and have a few weeks of data, then look at when your audience is actually around and shift your best posts into those windows. That is what best time to post on X is for - it is a second-order optimisation and should be treated as one.

Sustaining it

The reason most people fail at cadence is not that three posts a day is hard. It is that they set a volume they can only hit on a good week, miss it on a bad week, feel like they have broken something, and stop entirely.

A few things that actually help.

• Set the floor, not the target. Pick the volume you can hit on your worst week - maybe one post and ten replies. That is the commitment. Everything above it is a good day, not a requirement.
• Keep a running notes file. The bottleneck is rarely writing, it is remembering what you had to say. Capture the thought when it happens; write from the list, not from a blank box.
• Batch drafting, not sending. Write several posts in one sitting when you have momentum and send them across the day. You keep the human judgement and lose the daily blank-page tax.
• Timebox the replies. Thirty to forty minutes, a defined list of accounts, then close the app. Open-ended scrolling is what actually burns people out, not the writing.
• Measure weekly, not daily. Daily numbers are noise and checking them constantly is corrosive. Look at the trend once a week with the funnel calculator and get on with it.

Common questions

How many times a day should I post on X?

Two to three posts a day for an account trying to grow, alongside fifteen to twenty five replies. The posts give the algorithm more chances to find an audience for you; the replies are what actually put you in front of non-followers at a small follower count.

Can you post too much on X?

You can post too much of the same thing. Repeating one format all day teaches your followers to scroll past you, which weakens the early engagement signal every future post depends on. Ten varied posts is fine; ten near-identical ones is not.

Should I repeat a post format that worked?

Repeat the underlying quality, not the template. A format worked partly because it was fresh, and the third copy lands noticeably worse than the first. Take the lesson - it was specific, or contrarian, or useful - and express it in a different shape.

Is it better to post consistently or post a lot?

Consistently. Sustained daily activity compounds because it accumulates surface area and keeps you familiar to the same people; a burst followed by two silent weeks does neither. Pick a volume you can hold on a bad week and treat that as the floor.

Where Argus fits

Cadence dies when every session starts with "what do I even say". Argus removes the staring: it reads your own analytics to show what has been working, surfaces the live posts worth replying to right now, and drafts in your voice so you are editing instead of inventing. You always press send. Nothing is ever posted automatically.