Best Time to Post Methodology
TimeToPost publishes best-time-to-post curves from anonymized aggregate data where the sample is large enough. When a platform has not reached the public threshold, the guide is clearly labeled as an editorial baseline and updates as the dataset grows.
Based on editorial baselines, updating with real data as volume grows. Public aggregate platform data appears here once a platform reaches 50 analyzed posts.
Current Dataset Status
No public platform currently meets the 50-post threshold with engagement signal. The public pages use editorial baselines until enough anonymized TimeToPost data exists.
| Platform | Source used | Posts analyzed | Last updated |
|---|---|---|---|
| Editorial baseline | 0 / 50 | not yet updated from first-party data | |
| TikTok | Editorial baseline | 0 / 50 | not yet updated from first-party data |
| X (Twitter) | Editorial baseline | 75 / 50 | September 5, 2026 |
| Editorial baseline | 0 / 50 | not yet updated from first-party data | |
| Editorial baseline | 0 / 50 | not yet updated from first-party data | |
| YouTube | Editorial baseline | 0 / 50 | not yet updated from first-party data |
| Editorial baseline | 0 / 50 | not yet updated from first-party data |
How Scores Are Computed
Each eligible post is assigned to its UTC publish hour. We compute a weighted engagement value using likes, comments, shares, and engagement rate, with comments, shares, and engagement rate weighted above likes. For each hour, we average the weighted engagement across posts in that bucket, then normalize the strongest hour to 100. Every other hour is shown as a relative score against that peak.
Sample-Size Rule
A platform needs at least 50 analyzed posts with usable publish-hour data and non-zero engagement signal before TimeToPost labels the public curve as first-party data. Below that line, the page says it is using an editorial baseline. We do not inflate sparse samples into a data claim.
Account-Diversity Rule
Post count alone isn't enough. A platform is only labeled first-party once it also draws on posts from a minimum number of distinct accounts and organizations, so a single heavy user can never make a slice look representative of TimeToPost's customer base on their own. Both bars have to clear at the same time; a platform that only clears one of them still shows the editorial baseline. We track how many distinct accounts sit behind each slice internally, but we never publish that count or any account identifier: it exists purely as a gate before we say "TimeToPost data."
Editorial Fallback
Editorial baselines are platform-specific relative curves built from broad industry posting research and TimeToPost's internal default scheduling logic. They are not presented as first-party data. Once a platform reaches the threshold, the public API switches that platform to aggregate TimeToPost data automatically.
Privacy
The public dataset exposes only aggregate counts, platform-level source flags, and relative hourly scores. It never exposes account names, organization IDs, post content, platform post IDs, or absolute engagement totals.
Update Cadence
The public API is cached for crawler and page-render performance, and the methodology page revalidates daily. Dataset stats report the latest aggregate metric update timestamp available in the public dataset.
Open Dataset & License
The public best-times API behind this hub, and the per-country and per-weekday rhythm research behind the country and day guides, are published under CC BY 4.0. Use it however you want, including commercially, as long as you credit TimeToPost with a link back to this page.
Download it as machine-readable JSON from github.com/MelMayssonOwen/posting-times-dataset: 11 countries with their posting-rhythm research, 7 platforms, 18 industries, 20 formats and 7 weekdays, plus a zero-dependency TypeScript client. The data files are generated from the same source this site renders from, so they cannot drift out of sync with the guides.
Citation
Posting-times dataset by TimeToPost (https://timetopost.co/best-time-to-post/), licensed under CC BY 4.0.
The dataset is honestly labeled at the row level: curves marked editorial are compiled from published industry research, not TimeToPost's own posts, until a platform slice clears both sample-size gates above. If you cite this data anywhere the distinction matters, please carry that label forward.
Programmatic access: the same public, no-auth API this hub reads from is also wrapped by the TimeToPost MCP server, open source, so an AI agent can query best-time-to-post data the same way this page does.