The best time to post on Facebook: four studies, seven hours apart, and the metric Facebook switched off

Post between midday and 1pm, Tuesday to Thursday. That is the only window the four largest 2026 studies can all be true in at the same time, and we would rather give you a narrow honest answer than a confident wrong one. The wider and more reliable finding is the day rather than the hour: all four agree that Tuesday to Thursday beats Monday, Friday and the weekend, and they agree on almost nothing else.

What the four studies actually found

Each column below is the study's own published answer, copied rather than adjusted. The second column is ours: where we read the four as compatible. It is an inference and nobody measured it.

Seven hours apart on the same day

Look at the Tuesday row and then stop looking at anything else for a moment. Hootsuite's single best slot on Facebook is Tuesday at 5am. Sprout Social's Tuesday window is midday to 8pm. Neither is a fringe source. Both analysed over a million data points. They do not overlap by a single minute.

Monday is worse. Buffer puts Monday's three best hours at 9pm, 10pm and 7pm, an evening. RecurPost puts Monday at 9am to 12pm, a morning. Sprout Social puts it at midday to 1pm. Hootsuite says 5pm. Four studies, four different parts of the day, no two of them touching.

If you have read three of these guides and come away with three different hours, nothing is wrong with your reading. The field genuinely does not agree, and most articles on this search resolve that by picking whichever study they like and not mentioning the others.

The metric Facebook switched off

Here is the part almost nobody covering this keyword mentions, and it is the reason the disagreement above is not going to resolve itself. For years, the reliable answer to when to post was not a study at all. It was a chart inside your own Page insights showing when your followers were online, drawn from two Facebook API metrics: `page_fans_online` and `page_fans_online_per_day`.

Both were deprecated. According to trade reporting at the time, the deprecated metrics ceased to function on 16 September 2024, and calling them through the API returns an invalid metric error. Metricool, which builds its recommendations on that API, states that Meta stopped supporting best times on 24 September 2024, and that accounts connected after that date now see times derived from general behavioural studies rather than from the account itself.

That is the whole story of this keyword in one paragraph. The personalised answer went away, so every tool fell back to a pooled median across its entire customer base, and pooled medians across different customer bases produce different hours. The studies are not sloppy. They are measuring different crowds because the only thing that could have measured your crowd was switched off.

Why four serious studies still cannot agree

Beyond the missing metric, the four are not measuring the same thing, and stacking them in one table is unfair to all of them in the same way. It is worth knowing which is which before you weight one over another.

• Sprout Social counts engagements, nearly 2 billion of them across roughly 307,000 profiles. A study that counts engagements is partly measuring when large accounts post, because large accounts generate most engagements.
• Buffer counts posts, 14 million of them, and reports a median engagement rate. That treats a Page with 200 followers and one with 200,000 as one observation each, which is a different question and arguably the more useful one for a small business.
• RecurPost reports from scheduled publishing data, more than 2 million posts. That is a measure of when its customers chose to publish at least as much as when audiences responded.
• Hootsuite localised across 118 countries with the agency Critical Truth. Localising across 118 countries and reporting one clock is a defensible method that produces a number no single country experiences.
• All four measure people who pay for a scheduling tool, which is a more organised, more consistent, more marketing led account than the average small business Page. Nobody has the data for the other kind.

What to actually do on Monday morning

The useful version of all this is short, and most of it is not about the clock.

• Publish Tuesday to Thursday. This is the one thing all four studies agree on and the only part of this article we would defend strongly.
• Pick midday to 1pm and stop deliberating. It is the window with the most support across the four, and any hour you actually keep beats a better hour you keep abandoning.
• Check your own Page insights anyway. The API metric is gone but Meta still shows Page and audience activity inside its own tools, and your own data beats a pooled median from strangers every time it exists.
• Hold the slot for eight weeks before judging it. Fewer than eight data points cannot separate a posting time from a good week, and almost everyone changes the slot before that.
• Spend the saved deliberation on the post. What you post about moves results more than the hour does, and so does how often you post.

For the same question on other platforms, TikTok has its own four study comparison and the studies there agree with each other rather more than these do, while YouTube's timing turns on a different mechanic entirely because a video keeps earning for weeks. Instagram is the worst of the four for disagreement: one of its major studies lands twelve hours away from the other three. If Facebook specifically is the channel that matters to you, why your posts may not be shareable is a more common problem than the hour, and the cover photo is the one image on the Page that everybody sees.