There is no single best time to post on Instagram, and the biggest datasets of 2026 prove it by contradicting each other: Buffer's 9.6-million-post study crowns Thursday 9 a.m., Later's 6-million-post study says 5 a.m., Metricool's 24.4-million-post study says 6-9 p.m., and Sprout Social's 2-billion-engagement dataset says weekday early afternoons. They cannot all be right about you. What the disagreement actually reveals is more useful than any single answer, and Instagram's own statements point the same way.
What each study actually found
| Study | Sample | Metric | Best times | Published |
|---|---|---|---|---|
| Buffer | 9.6M posts, 200k+ accounts, Jan 2024-Dec 2025 | Median engagement rate | Thu 9 a.m., Wed 12 p.m., Wed 6 p.m.; evenings 6-11 p.m. strong | Jul 2026 |
| Sprout Social | ~2B engagements, 307k profiles (brand customers) | Raw engagements | Mon 2-4 p.m., Tue 1-7 p.m., Wed 12-9 p.m., Thu 12-2 p.m. | Mar 2026 |
| Later | 6M+ feed posts, 975k Reels, 2024 | Engagement rate, audience-local time | 5 a.m. overall; Monday 12-6 a.m. peak | Apr 2026 |
| Hootsuite | 1M+ business posts, 118 countries | Not defined | Mon 3-9 p.m., Tue 5-8 a.m. and 3-7 p.m., Wed 5 p.m. | Jul 2026 |
| Metricool | 24.4M posts, 375k accounts | Not specified | 6-9 p.m., heaviest 8-9 p.m.; Reels views peak 12 a.m. | Jun 2026 |
Every row links the study itself, with its sample size as published. Two of the five never define their engagement metric, which matters when you compare them.
Why the studies disagree
The contradictions are not sloppiness. They are four different methodology choices producing four different answers:
- Different populations. Sprout measures its own customers, who skew brand and business accounts; Later and Buffer skew creators and small accounts. Brand audiences engage on work breaks; creator audiences scroll at night. That single difference explains most of the afternoon-versus-evening split.
- Different metrics. Buffer uses median engagement rate (likes + comments + shares + saves over followers); Sprout counts raw engagements, which weights big accounts; Hootsuite does not say.
- Different timezone handling. Later normalizes to your audience's local time; others normalize to the poster. A "5 a.m." in one system is not a "5 a.m." in another.
- Different windows of data. The date ranges span 2024 through early 2026, across ranking changes.
The honest reading: each study is a real average over a population you may not belong to. Even the studies say so. Buffer: "timing can't save a weak post." Later: "Take these times with a grain of salt… finding your unique posting time matters most."
What Instagram itself says about timing
Instagram has never published a best time to post. In its official ranking explainers, timing barely appears:
"Mundane details like posting time": that is how Adam Mosseri's official post Shedding More Light on How Instagram Works files recency, inside "information about the post," well below your relationship with the viewer and their activity.
The 2023 Instagram Ranking Explained post lists the ranked signals for Feed, Stories and Reels; posting time is not on any of the lists. And in January 2025 Mosseri named the current top three signals: watch time, likes and sends; timing again absent (Social Media Today's coverage quotes and links his video).
Recency does matter: Feed shows recent posts, so posting when your audience is awake beats posting when they sleep. But that is a floor, not a growth lever.
So when should you actually post?
The defensible advice, given the evidence above:
- Start from your own audience, not a benchmark. Instagram's professional dashboard shows when your followers are most active; that beats every study's average because it is your population, not someone else's.
- Use a study window as a starting default only. If your audience matches a study's population (creator-heavy: Buffer's evenings; brand-audience: Sprout's afternoons), start there and measure.
- Test at consistent quality. Timing effects are small next to content effects; every study above says some version of this in its own caveats.
- Then judge by reach per post, not likes. Mosseri's stated metrics to watch are average watch time, likes per reach and sends per reach (source).
Scheduling makes the testing part free: queue the same weekly content at different slots and compare. That is the workflow ad2app is built for, and the free tier includes a posting field guide for every platform, emailed when you sign up, the Instagram guide covers the full ranking picture behind this article.
Every claim above links its primary source. The per-platform specifics live in the free field guides, sources included.