Search for the best time to post on X and you will find a dozen articles quoting near-identical tables. Weekday mornings around nine. Lunchtime. A second peak in the evening. Tuesday and Wednesday better than Friday.
The tables are not wrong, exactly. They are averages of enormous datasets, they are almost always in US Eastern time, and they almost never say so. If you are in London writing for an American audience and you follow one literally, you are posting at 2pm your time and concluding that mornings are dead.
The timezone is the whole question
"Post at 9am" is not advice until you say whose 9am. The variable that matters is not the clock on your wall, it is whether the people you are writing for are awake and holding a phone.
For a European founder writing for American developers, the useful windows are the early afternoon and evening in Europe, because that is the American morning and lunchtime. For an Indian account writing for an Indian audience, the strongest window is often late evening, well after the American tables have gone quiet.
This is the specific thing our posting time planner does. You say where your audience is, it converts the windows into your local time, and it handles daylight saving properly, which means a slot can legitimately land on a different weekday than you expected.
How much does timing actually matter?
Less than the volume of articles about it would suggest, and it is worth being blunt about that.
Posting time is a tiebreaker. A mediocre post at the perfect hour still loses to a good post at an average hour. No schedule rescues writing nobody wanted to read, and the accounts that grow are not the ones that found a magic slot.
Where timing does earn its keep is on the posts that were already going to do reasonably well. Ranking on X responds to early engagement, and early engagement depends on there being people around when the post goes out. Publishing a good post into a dead window wastes it. Publishing it into a live one gives it the initial signal that compounds into reach.
So the honest framing is not that good timing makes posts succeed. It is that bad timing makes good posts fail, and that is worth avoiding.
Weekends are not the wasteland they are described as
General advice says weekend volume is lower. It is, but so is competition, and the ratio matters more than the absolute number.
For developer, founder and indie-hacker audiences, weekends are often genuinely strong. That crowd uses X recreationally rather than professionally, which means Saturday morning is leisure time rather than work time, and leisure time is when people read long things.
If your audience is corporate or B2B, the standard advice holds better: weekday business hours, with Friday afternoon dropping off sharply.
Replies run on a different clock entirely
If most of your growth comes from replying rather than posting, everything above is the wrong frame.
A reply is not competing for a slot in a timeline. It is competing for attention inside somebody else's conversation, and that conversation has a lifespan measured in tens of minutes. What matters is not being active at 8am, it is being early to a post that is gaining traction.
The first few replies on a post that goes on to do well get a disproportionate share of its attention. The fortieth reply on the same post is invisible. That is a monitoring problem, not a scheduling one, and it is why "reply consistently" is much harder advice to follow than "post consistently".
Doing it by hand means checking X constantly, which is the failure mode most people fall into. ReplyBoss watches the accounts and keywords you choose and surfaces conversations while they are still early, which is the part a calendar cannot help with.
Finding your own numbers
Use published windows as a hypothesis, then check them against your own data. X's analytics show impressions per post with timestamps, which is enough.
Post at varied times for a month, then group the results by slot. Two cautions when you read it back. Do not compare a Tuesday thread against a Sunday one-liner and conclude Tuesday wins, because you have changed two variables. And give every slot at least four or five posts before trusting it, because individual posts vary far more than posting times do.
What usually emerges is narrower than the published tables suggest and specific to your audience. That is the number worth building a schedule around.
A reasonable default
If you have no data yet and want somewhere to start: pick the two windows when your audience is most likely awake and unoccupied, post consistently into those for a month, and spend the effort you would have spent optimising the schedule on replying instead.
The planner will give you those two windows in your own timezone. It is a starting hypothesis, not a rule, and the fastest way to improve on it is your own analytics.