Best Time to Post on Twitter
Choose an audience timezone for general guidance. Add any public @handle when you want a personal map built from returned original posts in the last 90 days.
Choose a timezone or analyze a public Twitter account
General Twitter posting windows in audience-local time.
Use this directional baseline when no handle is provided. Add a public handle above to replace it with the account's returned 90-day post performance.
| Day | 12a | 3a | 6a | 9a | 12p | 3p | 6p | 9p | ||||||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Mon | ||||||||||||||||||||||||
| Tue | ||||||||||||||||||||||||
| Wed | ||||||||||||||||||||||||
| Thu | ||||||||||||||||||||||||
| Fri | ||||||||||||||||||||||||
| Sat | ||||||||||||||||||||||||
| Sun |
General peak
General moderate
No highlighted window
Times are interpreted in the audience timezone selected above.
General windows
A strong cell says what happened—not why.
Timing can help a good post meet an active audience, but public data cannot isolate timing from topic, format, news cycles, or one unusually large result.
Shown here
The general view shows published audience-local windows. Add a handle to measure posting day and hour, likes, replies, reposts, returned public views, sample size, and the selected timezone.
Not inferred
Follower online activity, audience location, private impressions, link clicks, conversions, or a guaranteed future posting time.
How the calculation works
The general view is published guidance. Personal results show the exact window, returned-post count, timezone, and low-confidence cells used in the calculation.
- 01
Choose the general or personal view
Leave the handle blank for audience-local guidance based on Sprout Social's 2026 X/Twitter benchmark. Add a handle and CarryFeed instead requests recent public timeline pages, removes replies and reposts, and stops at 200 returned posts or the 90-day boundary.
- 02
Use the audience's local clock
General windows are read directly in the selected audience timezone. For a personal result, each returned post timestamp is converted into that timezone. Changing it never spends another analysis.
- 03
Compare robust middle values
Core interactions are likes + replies + reposts. Each hourly cell uses its median and is pulled toward the account-wide median when its sample is thin, reducing the influence of a one-off spike.
- 04
Expose weak evidence
Every populated cell reveals how many posts support it. Missing public view counts stay missing, and reports with fewer than 10 posts remain visibly low-confidence.
Use two windows, then test them with comparable posts.
Pick one repeated high-performing window and one practical alternative. Post comparable formats in both for a few weeks, then rerun the same 90-day view. A heatmap becomes useful when it changes a repeatable decision, not when it produces one “magic hour.”
Prefer repeated cells. Three ordinary posts are usually a better scheduling clue than one viral result.
Match the audience clock. If readers are split between regions, recast the same report twice and look for overlap.
Keep content type comparable. A launch thread and a casual reply do not make a clean timing test.
Questions before you schedule.
Use the report as a repeatable test, not a promise that one hour will always win.
What happens if I leave the Twitter handle blank?
CarryFeed shows general audience-local posting windows in the timezone you selected. This view is directional platform guidance, not a claim about one account's followers or performance, and it does not run a personal account analysis.
What time range does this Twitter posting-time tool analyze?
When you add a handle, CarryFeed uses returned original public posts published during the last 90 days, up to 200 posts. The result shows the exact dates and post count that were actually available, so a shorter or partial sample is never presented as complete 90-day history.
Does the heatmap show when my followers are online?
No. The general view shows published audience-local windows, while a personal view shows how the account's returned public posts performed at each posting time. Neither view can see private follower activity, audience location, impressions, or the account owner's X Analytics.
Should I choose my timezone or my audience's timezone?
Choose the timezone where the audience you want to reach is most likely to read. The general windows are interpreted in that local time; a personal result recasts the same returned post timestamps. Changing the selector does not run or charge for another analysis.
How does CarryFeed rank the best hours?
The general view labels published peak and moderate windows rather than inventing account scores. In a personal view, CarryFeed compares median likes + replies + reposts for posts in each day-and-hour cell. A separate interaction-rate view uses returned public views when they exist, and thin cells are pulled toward the account-wide median so one viral post has less influence.
How many posts do I need for a useful recommendation?
Ten posts is a bare minimum for orientation, not certainty. Repeated posts in the same window are more useful than a single result. CarryFeed labels single-post cells and low-coverage reports instead of replacing them with an unrelated platform average.
Can I analyze someone else's Twitter account?
Yes, if the account and its posts are publicly available. Private, deleted, restricted, suspended, or otherwise unavailable activity cannot be analyzed.