Three recommended windows and two best days for your niche, converted to your timezone. Heuristic starting points — you refine them by posting and measuring.
Maker audiences check X at US morning coffee, lunch, and after work.
These are heuristic starting points based on where each niche's audience concentrates (mostly US-centric attention patterns), not measurements of your account. Post in these windows for two weeks, compare views and engagement, then adjust. Your own data beats any generic chart.
The "For You" timeline was supposed to make posting time irrelevant — the algorithm surfaces good posts whenever they were published, right? Partly. What actually happens is that the algorithm makes an early judgment call: in the first hour or two, it shows your post to a slice of your followers and watches. Strong early engagement earns wider distribution, including to non-followers; a quiet first hour usually means the post is done, regardless of its quality. That early test is why timing persists as a lever. A post published while your audience sleeps takes its algorithm exam in an empty room. The same post published into your audience's morning scroll gets its likes and replies while the test is still running.
We'll be direct: there is no public dataset of X engagement by niche and hour, and any tool claiming precise, study-backed numbers is overselling. What exists is a set of durable attention patterns that practitioners keep rediscovering. English-language X attention is heavily US-weighted, so windows cluster around US Eastern rhythms: the 8–10am coffee scroll, the lunch dip, the 5–8pm decompression block, and a late-night tail that's strong for developers and fitness content. Niches shift within that frame — finance attention tracks market hours, B2B marketing lives midweek during work hours, essays get read on slow Sunday mornings. The tool encodes those patterns as two-hour windows and converts them to your clock. They're priors, not answers — which is exactly how you should use them.
Here's the process that turns a generic heuristic into your personal schedule. Pick two of the three recommended windows. For two weeks, alternate your main daily post between them, holding everything else as constant as you can — similar effort, similar formats, no cheating by saving your best material for your favorite slot. After ten or more posts per window, compare the median views and engagement per window; medians matter because one semi-viral post will otherwise decide the experiment. Keep the winning window as your incumbent and test it against a third challenger. Two or three cycles of this — about six weeks — gives you a schedule derived from your actual audience, and it will sometimes contradict every chart on the internet. Trust it anyway.
A slightly suboptimal window hit every single day beats the perfect window hit twice a week. Part of this is audience training — regulars start expecting you, and their reliable early engagement is exactly the signal the algorithm's first-hour test wants to see. Part of it is volume: more posts means more lottery tickets and, more importantly, more data for your calibration loop. This is also the honest case for scheduling. If your audience's prime window is 3am your time, the answer isn't heroic sleep deprivation; it's writing when you're sharp and scheduling into the window — then showing up in the replies when you wake up, because the conversation half of the game can't be scheduled.
Timing advice applies to evergreen posts — the opinions, lessons, and stories that make up most of a content strategy. It does not apply when news breaks in your niche: reacting well in the first thirty minutes of a story beats posting at the "right" hour by a mile, because you're riding a live attention wave. It also matters less for replies, which inherit the timing of the conversation they join. So hold the windows loosely: they're for the 80% of your posting that's planned, and permission to break the schedule is built into the other 20%.
No, and we label them accordingly. They're heuristic starting points built from where each niche's audience concentrates — mostly US-centric attention rhythms like morning commutes, lunch breaks, market hours, and evenings. Nobody outside X has post-level timing data for your specific followers. The honest workflow is: start with a sensible heuristic, post consistently, measure, adjust.
Less than it used to, but it still matters. The algorithm decides distribution largely from how a post performs in its first hour or two — early likes and replies expand reach, silence buries it. Posting when your audience is awake and scrolling maximizes that early signal. Time won't save a weak post, but it can noticeably amplify or handicap a decent one.
Your audience's, always. If you're in Karachi or Berlin writing for a US-heavy indie hacker audience, your best windows may be your late evening. The timezone selector in this tool converts the niche's windows into your local clock so you know when to actually be at the keyboard — or when to schedule.
For most accounts, 1–3 quality posts beats 5+ mediocre ones — each weak post trains both the algorithm and your followers to expect less. Use one window for your main post and the others for replies and engagement, which matter as much as posting for early-stage accounts.
Run a simple two-week experiment: pick two of the recommended windows and alternate posts between them, keeping format and effort roughly constant. After 10+ posts per window, compare median views and engagement (medians resist outliers). Keep the winner, test it against a new challenger window. Two or three cycles of this beats any generic chart.
Directionally, yes. B2B and marketing content does best midweek when people are in work mode; essays and reflective writing travel well on Sundays; fitness content spikes Mondays (resolution energy) and weekends. Treat the day recommendations like the hour recommendations: reasonable priors to verify against your own numbers.
DeckPost drafts in your voice, schedules at your peak windows, and reviews the numbers with you every week.