There is no single best time to post on Instagram. That is not a dodge. It is what the data actually says once you put five major studies side by side. Buffer’s 9.6 million posts point to Wednesday evening. Sprout’s two billion engagements point to Tuesday and Wednesday afternoons. Hootsuite points to Monday, Tuesday, and Thursday. Later points to 5 a.m. Metricool points to 8 p.m. and measures views, not engagement, so its answer is not even the same kind of answer. The honest short version: midweek evenings are where the most studies overlap, Wednesday evening most of all, and every single study says your own Insights data beats all of them. Start with Wednesday evening, then test it against your own numbers.
All five of these companies sell scheduling software, which is worth remembering when reading their data. We compare what they actually cost, and what their free plans really give you, in our SMM tool comparison.
Key takeaways
- No study in this guide declares one universal best time, and neither does this guide. Every number below comes with its sample, its date window, and its metric.
- Four of the five studies agree on two things: midweek is strong, and Saturday is weak.
- The biggest split is early mornings. Later says 5 a.m. is the best time. Buffer, Sprout, and Hootsuite say early mornings are dead or weak.
- The overlap window, computed from the five studies, is Wednesday evening, roughly 5–9 p.m. in your audience’s local time. That is this guide’s synthesis, not any vendor’s claim, and not a guarantee.
- Instagram itself publishes no universal best time. Its documented route is your own Insights “Most Active Times,” and all five vendors say that overrides their benchmarks.
Jump to a section
- The five studies, side by side
- Where 4 of 5 agree
- Where they split: the 5 a.m. paradox
- Why the numbers disagree
- The overlap window to test first
- Which study matches your account
- The 4-week two-slot test
- How to find your own best time in Insights
- What matters more than timing
- Frequently asked questions
What do the 5 studies say, side by side?
Read the sample and metric columns before the best-time column. A study of 9.6 million Instagram posts and a study of views across professional accounts are not measuring the same thing, and their answers should not be compared as if they were. All times below are in the audience’s local time, except Metricool, whose timezone handling is not stated on its pages.
| Study | Sample | Date window | Metric | Best day | Best window | Worst day | Timezone |
|---|---|---|---|---|---|---|---|
| Buffer | 9.6M Instagram posts, 200K+ accounts | Jan 2024–Dec 2025 | Median engagement rate per follower (formula stated) | Wednesday | Thu 9 a.m., Wed 12 p.m., Wed 6 p.m.; Tue 7 p.m. (Tuesday best) | Friday, Saturday | Normalized to local time |
| Sprout Social | ~2B engagements, ~307K profiles, six networks (not Instagram-only) | Nov 2025–Feb 2026 | Peak engagement windows (formula not stated) | Tuesday, Wednesday | Tue 1–7 p.m., Wed 12–9 p.m. | Saturday (lowest) | Local time |
| Hootsuite | 1M+ posts with Critical Truth (cross-platform wording) | 2025 analysis; page updated Oct 2026 | Engagement peaks (formula not stated) | Monday, Tuesday, Thursday | Mon 3–9 p.m., Tue 5–8 a.m. + 3–7 p.m., Wed 5 p.m., Thu 4–5 p.m. | No worst day named (weakest before 8 a.m., after 10 p.m.) | Localized across 118 countries |
| Later | 6M+ Instagram feed posts + 975K Reels | Jan–Oct 2024 (oldest dataset) | Engagement rates over time (formula not stated) | Monday | 5 a.m. local; 3–6 a.m. above average daily | Saturday | Normalized to local time |
| Metricool | 24.36M Instagram posts, 375,118 accounts | Jan–Feb 2025 + Jan–Feb 2026 | Views (not engagement) | Wednesday (Friday and Tuesday close behind) | 8 p.m. peak (Tue and Sun at-a-glance), 7–9 p.m. window | Saturday | Not stated |
Two rows need a warning before you use them. Hootsuite’s page title says “2026 data,” but its body repeatedly calls the work its 2025 analysis, and its timezone heatmaps are labelled Q4 2024. The honest description is an analysis published July 2026 and updated October 2026, based on its 2025 analysis. Metricool’s own pages contradict each other: one line calls 6–9 a.m. a sweet spot while another says 8 p.m. has the most views and the strategy guide says 7–9 p.m. This guide treats Metricool’s answer as the evening, 7–9 p.m. peaking at 8 p.m., and shows the morning line beside it rather than hiding it. Note: some roundups quote Metricool as 6–9 p.m.; Metricool’s own pages say 7–9 p.m. (strategy guide) with an 8 p.m. peak.
Where 4 of 5 agree
Midweek is strong. Tuesday through Thursday appears in the best days of Buffer (Wednesday, followed by Thursday and Tuesday), Sprout (Tuesday, Wednesday), and Hootsuite (Monday, Tuesday, Thursday); Metricool’s leading day is Wednesday, with Friday and Tuesday close behind. Later is the exception on best day, naming Monday, but its 3–6 a.m. window applies to every day including midweek.
Saturday is weak. Buffer names Friday and Saturday as worst days, Sprout names Saturday the single lowest day, Later names Saturday the worst day overall, and Metricool names Saturday worst, followed by Sunday and Thursday. Four of five agree, and that is about as close to consensus as this topic gets.
Your own data wins. All five vendors say it in their own words: check your audience’s activity in Insights and trust that over any benchmark, including theirs. That is the one unanimous finding in the whole set, and it is the conclusion of this guide too.
One more partial agreement, stated carefully because Later breaks it: the overnight dead zone. Buffer calls 1–5 a.m. the worst window, Sprout tells readers to avoid the 3–7 a.m. dead zone, and Hootsuite calls before 8 a.m. among the least effective times. Three of five say overnight is dead. Later says the opposite, which is the next section.
Where they split: the 5 a.m. paradox
Later’s headline finding is 5 a.m. in the audience’s local timezone, with 3–6 a.m. above average every single day. Buffer says weekday mornings from 6–11 a.m. are generally weak and 1–5 a.m. is the worst window on any day. Sprout says avoid 3–7 a.m. on workweek days. Hootsuite says before 8 a.m. is among the least effective times. These cannot all be right about the same audience, and this guide does not pretend they are.
Later gives its own explanation, and it deserves to be heard alongside the contradiction rather than instead of it: fewer posts compete in the early hours, and a post published early accumulates engagement as users do their first scroll of the day. That mechanism is plausible for Later’s creator-leaning user base, and it is exactly why a benchmark built on one platform’s users does not transfer cleanly to another’s. Later also says it plainly on its own page: take these times with a grain of salt, they are a starting point, not a final answer, and if your analytics show your audience engages most at 2 p.m., trust that over any study, including theirs.
The disagreement is the useful part of this topic. A single “best time” chart would be easier to publish and easier to click, and it would be wrong for a large share of readers. The split tells you something real: posting time interacts with who your audience is and when they actually open the app, so a benchmark can only ever be a starting window to test. Buffer puts it better than I can: take these numbers as guidance, not gospel, and timing can’t save a weak post. Think of it as the cherry on top of your strategy, not the cake itself.
Reels timing splits three ways too, and the dataset notes matter. Later’s Reels answer is 12 a.m. on Monday, from a Reels-specific dataset of 975,000 Reels. Buffer’s Reels note is evenings 6–11 p.m. on Wednesday and Thursday, but Buffer analyzed all formats together, so that is not a Reels-only finding. Sprout says Reels times mirror the overall windows because the algorithm prioritizes when your audience is active over content format. Three answers, three different dataset notes. Do not pick one and drop the notes.
Why the numbers disagree
Different measures first. Buffer states a formula: median engagement rate equals likes plus comments plus shares plus saves, divided by followers, times 100, with the median used so outlier posts and huge accounts do not skew the result. Sprout, Hootsuite, and Later do not state an engagement formula on their pages, so their “engagement” windows cannot be compared to Buffer’s rate as if they were the same metric. Metricool measures views, full stop. A post can rack up views at 8 p.m. without earning the engagement Buffer counts at 6 p.m., and both findings can be true at once.
Different user bases second. Buffer’s sample is 200,000+ mixed accounts. Sprout’s comes from 30,000+ customers, whose roughly 307,000 profiles were analyzed, and leans brand and business. Hootsuite’s spans 118 countries. Later’s leans creator. Metricool’s skews professional accounts with a Spain and Latin America lean. A best time for one base is not a best time for another, which is also why Sprout’s industry splits vary so much: its page puts Retail at Monday to Friday 12–3 p.m. as one example, while other industries peak elsewhere. If your niche has its own row on a vendor page, use that row over the global window.
Different windows third, and this is where dating every row earns its keep. Later’s data is January to October 2024, the oldest in the set, published April 2026. Buffer’s runs January 2024 to December 2025. Sprout’s is the newest and shortest, November 2025 to February 2026. Metricool compares January–February 2025 with January–February 2026. Hootsuite’s window is the least precise, described only as its 2025 analysis. Instagram changed a lot between 2024 and 2026, so an undated number in a roundup is not just incomplete, it is misleading.
The overlap window to test first
Computed from the table above, Wednesday evening, roughly 5–9 p.m. in your audience’s local time, is the single window where four studies overlap: Buffer’s Wednesday 6 p.m. slot, Sprout’s Wednesday 12–9 p.m. window, Hootsuite’s Wednesday 5 p.m., and Metricool’s Wednesday leading day with its 8 p.m. peak. Tuesday evening is the second overlap: Buffer’s Tuesday 7 p.m., Sprout’s Tuesday 1–7 p.m., Hootsuite’s Tuesday 3–7 p.m., and Metricool’s Tuesday close behind with its 8 p.m. peak.
Label this correctly in your head. The overlap window is this guide’s synthesis from the five studies. No vendor claims it, and it is not a guarantee of anything. Later’s 5 a.m. finding does not overlap with either window, which is the point: if your audience genuinely scrolls at dawn, the overlap window will lose to Later’s window for you, and your Insights will show it. Treat Wednesday evening as a sensible starting window to test, nothing stronger.
Which study matches your account
This is editorial guidance for picking a starting point, not data, and each match carries its caveat. Creator-heavy small accounts are closest to Later’s dataset, which is Instagram-specific and creator-leaning, but it is also the oldest data here. Brand and business accounts are closest to Sprout’s customer base, but Sprout’s headline figure covers six networks, not Instagram alone. Mixed small-business accounts are closest to Buffer’s 200,000+ mixed accounts, and Buffer is the only study that states its formula. Multi-country audiences are closest to Hootsuite’s 118-country localization, but its data window is the least precise. Professional accounts with a Spain or Latin America lean are closest to Metricool’s base, but Metricool measures views, not engagement, and its timezone handling is unstated. Pick the closest match as your second test slot, not as your answer.
Later also splits by follower count, which no other study in this set does, and it is worth one line: nano accounts (1,000–5,000 followers) peak Monday to Friday at 5 p.m., micro accounts (5,000–25,000) at Monday 11 p.m. and Friday 12 a.m., mid-tier (25,000–100,000) at Sunday 12 a.m. and Monday 11 p.m., macro (100,000–500,000) at Sunday and Monday 12 a.m., and mega accounts (500,000+) at Monday 10 a.m. The page does not state sample sizes per tier, so treat the tier table as directional, not as robust as Later’s headline 6 million finding.
The 4-week two-slot test
- Pick two slots: the overlap window (Wednesday evening) and one vendor window that fits your audience from the section above.
- Post the same content type in both slots for 4 weeks. Do not compare a Reel in one slot with a carousel in the other, or the format difference will swamp the timing difference.
- Compare median engagement per follower in Insights across the two slots, not your single best post in either slot. One viral post is noise; the median is the signal, which is why Buffer uses it.
- Keep the winner and re-test once a quarter. Audiences shift, and Hootsuite notes holidays, school schedules, and current events move behavior around.
- Post consistently while you test. Buffer suggests 3–5 Instagram posts per week plus 1–2 Stories daily, and Hootsuite suggests 3–5 posts per week with 2 Stories per day. A test with two posts total proves nothing.
How to find your own best time in Insights
Every vendor describes the same path, and this guide attributes the steps to their guides rather than to a Help Center page, because no Help Center page stating a universal best time was found in the research behind this guide. On a professional (Business or Creator) account, open the Professional Dashboard, go to Total Followers, and open Most Active Times, which shows activity by day and by hour. That view is your audience’s actual behavior, and it is the number every benchmark in this guide defers to. Check it by day first, then by hour: a strong day with a weak hour inside it is still a weak slot.
What matters more than timing
Buffer, citing Head of Instagram Adam Mosseri, reports two things worth more than any slot in the table. First, Instagram’s most important signal in 2026 is sends per reach, how often people send a post or Reel to a friend. Second, Mosseri’s timing advice as reported by Buffer is to post 30 to 60 minutes before your audience is likely to come online, so the post is waiting when they open the app. Both points are Buffer’s reporting of Mosseri, not pages on instagram.com read for this guide, and this guide keeps that attribution attached. The practical read is simple: a post people want to send beats a post published at the perfect hour, and timing works best as the last adjustment to a strong post, not as a rescue for a weak one.
Frequently asked questions
Is there one best time to post on Instagram?
No. Five major studies give five different answers, and they measure different things across different audiences and date windows. The closest thing to a consensus is midweek strength, Saturday weakness, and Wednesday evening as the window where four studies overlap. Your own Insights data overrides all of them.
Why does Later say 5 a.m. when everyone else says mornings are dead?
Later’s dataset is Instagram-specific and creator-leaning, and its explanation is less feed competition plus the first-scroll-of-the-day habit: posts published early accumulate engagement as users wake up. Buffer, Sprout, and Hootsuite measure different user bases and find early mornings weak or dead. Later’s data is also the oldest here (January–October 2024). Test the 5 a.m. window against your own audience rather than accepting or dismissing it.
Does Metricool measure the same thing as Buffer?
No. Metricool’s best time is measured by views, while Buffer measures median engagement rate per follower with a stated formula. A time can win on views and lose on engagement, which is one reason Metricool’s 8 p.m. peak differs from Buffer’s Wednesday 6 p.m. slot. Compare the metric before comparing the hour.
Why is Hootsuite’s page titled 2026 data if it cites a 2025 analysis?
Because the page was published in July 2026 and updated in October 2026, while its body repeatedly calls the work its 2025 analysis and its timezone heatmaps are labelled Q4 2024. This guide describes it honestly as an analysis published July 2026 and updated October 2026, based on its 2025 analysis, and does not claim a 2026 data window for it.
Should small accounts follow the same times as big accounts?
Only Later splits its headline windows by follower count in this set, and it does not state sample sizes per tier, so treat its tier table as directional. The safer route is the 4-week two-slot test above: run the overlap window against one vendor window and let your own median engagement per follower decide.
Testing a posting window properly means scheduling consistently for a few weeks, which is easier with a free scheduling tool than by hand.
Do Reels need a different posting time than photos?
The studies disagree, and their dataset notes explain why. Later’s Reels answer (12 a.m. Monday) comes from a Reels-specific dataset of 975,000 Reels. Buffer’s Reels note (evenings 6–11 p.m., Wednesday and Thursday) comes from all formats pooled together, not a Reels-only split. Sprout says Reels times mirror the overall windows because audience activity matters more than format. Do not quote any of the three without its dataset note.
What about Stories? Is there a best time for them?
No vendor in this set publishes a Stories-specific dataset. Later’s Stories hours (early morning 6–8 a.m. and late evening 7–9 p.m.) are framed on its page as guidance from research and industry patterns, not dataset output. Buffer’s Stories advice is consistency (1–2 daily) over slot-picking. Treat any Stories hour as guidance, not data.
Sources
- Buffer: Best Time to Post on Instagram, Data from 9.6M Posts (updated September 2026)
- Sprout Social: Best Times to Post on Instagram in 2026
- Hootsuite: The best time to post on Instagram (published July 2026, updated October 2026)
- Later: Best Time to Post on Instagram (published April 2026, data January–October 2024)
- Metricool: Best Time to Post on Social Media in 2026
- Metricool: 2026 Instagram Marketing Strategy Guide (7–9 p.m. window)
- Social Media Today press release: Metricool’s 2026 Instagram Study methodology (sample and date windows)
