x3 studies plotted, 3 read
Best time to post on X
Mid-morning wins on X, and the largest study finds it wins on every single day of the week - including both weekend days. This is the platform where the day matters least.
The week, by how many studies agree
Each cell is one hour. The darker it is, the more of the 3 studies we plot name that hour for X. It counts agreement between studies, not engagement - these samples measured different things and cannot be averaged.
Wednesday, 9am to 10am
2 of 3 studies name this hour: Buffer, Hootsuite
| Day | 12am | 1am | 2am | 3am | 4am | 5am | 6am | 7am | 8am | 9am | 10am | 11am | 12pm | 1pm | 2pm | 3pm | 4pm | 5pm | 6pm | 7pm | 8pm | 9pm | 10pm | 11pm |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Mon | ||||||||||||||||||||||||
| Tue | ||||||||||||||||||||||||
| Wed | ||||||||||||||||||||||||
| Thu | ||||||||||||||||||||||||
| Fri | ||||||||||||||||||||||||
| Sat | ||||||||||||||||||||||||
| Sun |
| Day | Best window | Studies agreeing |
|---|---|---|
| Monday | 8am-11am | 1 of 3 |
| Tuesday | 9am-6pm | 1 of 3 |
| Wednesday | 9am-11am | 2 of 3 |
| Thursday | 9am-11am | 2 of 3 |
| Friday | 9am-11am | 2 of 3 |
| Saturday | 9am-12pm | 1 of 3 |
| Sunday | 8am-11am | 1 of 3 |
What is specific to X
- Buffer's X data is the flattest result in this entire dataset: 9am to 11am ranks top on all seven days, Saturday and Sunday included. Across 8.7 million posts they found almost no day-of-week signal at all. That is a finding, not a gap - on X, the hour matters and the day mostly does not.
- Hootsuite frames the mid-morning peak by what X is for: "X is a news-driven platform, so mid-mornings hit the sweet spot". Posts here have the shortest useful life of any platform on this site, which is what makes the hour matter more than the day.
- We plot only one Hootsuite window for X. Their dedicated X page redirected to an image rather than an article on the day we checked, so we take the single line their round-up publishes and claim nothing beyond it.
Where the studies contradict each other
Every other page on this subject prints one confident grid. These are the places our sources flatly disagree, with both sides named, because that disagreement is the most useful thing we found.
Morning or afternoon?
- Buffer9am to 11am, top-ranked on every day of the week.
- Sprout Social"Tuesdays, Wednesdays and Thursdays 12-6 p.m." - an afternoon window.
Buffer and Hootsuite both land on mid-morning; Sprout is the one study pulling toward the afternoon. Two against one is as close to consensus as this category gets.
Does the weekend work on X?
- BufferRanks Saturday 9am to 11am identically to the weekdays, though it names Saturday and Friday its worst days overall.
- Sprout SocialPublishes no weekend hours and calls weekends its worst days.
Buffer's own numbers undercut its own worst-day label here. If the mid-morning hour holds every day, the weekend penalty on X is small.
Why this matters less than the rest of the page suggests
Every study on this page is published by a company that sells scheduling software, and most of them tell you plainly that timing is a small lever. We think that is the single most useful thing in the whole category, so we put it here rather than at the bottom.
Only one study we found measured the size of the effect instead of just ranking hours. AuthoredUp, across more than 3 million LinkedIn posts, found the gap between the best and worst day is under 10% for personal profiles, and that weekday against weekend engagement lands within roughly 7%. Company Pages see a larger swing, about 1.7 times between Wednesday and Sunday.
Set against that, the studies here disagree with each other by up to sixteen hours on the same platform. When the disagreement between sources is far larger than the effect any of them measures, precision to the hour is false precision.
The honest version: post in the broad window your audience is awake and reachable, keep the cadence steady, and spend the effort you saved on the post itself. Then replace all of this with your own analytics as soon as you have enough of your own posts to read.
Questions
- What is the best time to post on X (Twitter)?
- Mid-morning, roughly 9am to 11am. Buffer found that window top-ranked on all seven days across 8.7 million posts, and Hootsuite independently names Wednesday to Friday 9am to 11am. Sprout is the dissenter, preferring midday to 6pm on Tuesday through Thursday.
- Does the day of the week matter on X?
- Barely, on the largest sample available. Buffer's day-by-day table is almost flat - the same mid-morning hours win Monday through Sunday. If you only change one thing about your X schedule, change the hour, not the day.
- How long does a post on X stay alive?
- Shorter than anywhere else covered here, which is why timing carries more weight on X than on Instagram or YouTube. Hootsuite attributes the mid-morning peak to X being a news-driven platform, where posts are read close to when they are published.
- Is it still worth posting on X at the weekend?
- Buffer names Saturday and Friday its worst days yet still ranks Saturday 9am to 11am alongside the weekdays, and Sprout publishes no weekend data at all rather than measuring it as poor. The evidence for a real weekend penalty on X is weak.
The studies, read 2026-08-25
Every window above comes from one of these pages. A study we could not put on a clock is listed here too, marked, rather than quietly dropped.
- Sprout SocialBest Times to Post on X (Twitter) in 2026Nearly 2 billion engagements across roughly 307,000 profiles, 27 Nov 2025 to 27 Feb 2026.Median engagement per hour across the profiles of 30,000+ Sprout customers.Reports in local time. Sprout's 2026 refresh dropped the old US Central Time hub: "schedule your posts for the hours listed below in their respective local time zones".Sprout names no Saturday or Sunday windows on any platform - it calls weekends its worst days and publishes no weekend hours. Read an empty weekend cell as not published, not as no engagement.page dated 2026-03-31we read it
- BufferThe Best Time to Post on X in 20268.7 million posts.The time slots with the highest median engagement.Normalised to the reader's local time, so no conversion is needed - Buffer says its data team makes the recommended timezones "universally applicable", without publishing the method.Buffer's X windows are almost flat: 9am to 11am wins every day of the week, weekends included. On this platform its data shows barely any day-of-week signal at all.page dated 2026-03-13we read it
- HootsuiteBest Time to Post on Social Media in 2025 (by platform)Over 1 million posts across industries and networks.Posts analysed with the data-science agency Critical Truth, localised across 118 countries.Normalised across 118 countries and stated as time-zone agnostic. The normalisation method is not published.Hootsuite's dedicated X page redirected to an image rather than an article on the day we checked, so we take the one window their round-up states for X and plot no more than that.page dated 2025-11-19we read it
How we built this
We read each study's own page, transcribed its day-by-day windows, and recorded the clock it publishes on. The heatmap shades each hour by how many independent studies name it. It is a count of agreement, not an engagement score: we never blend numbers from studies that measured different things in different ways.
A study only reaches the map if we can put its hours on a clock. Metricool measures the largest samples in this category and states no timezone on any page we read, so its hours are quoted in the text and left off the grid. HubSpot surveys marketers about what they believe rather than measuring posts, so it is left off too. Both are listed with the rest, with the reason attached.
Most publishers now normalise their windows to the reader's local time, so we plot those as published rather than converting them - converting an already-local window is how these pages usually go wrong. Later's TikTok windows are the exception, published in PST, so those we convert to your clock with daylight saving resolved for today.