Best Time to Post: Instagram, TikTok & YouTube Analyzer

When should this account post? Give this actor the post rows your Instagram, TikTok, YouTube, X or Facebook scraper already returned (a dataset, CSV or Google Sheet) and it ranks the weekday and hour slots when that account's posts did best, in your audience's time zone. Not a generic "post at 11am on Tuesday" blog chart: the answer comes from the account's own posts, or from your rivals' posts.

No scraping, no login, no AI fees: it is maths on rows you already have, so it runs in seconds for a few cents.

Who uses it

• Social media managers and agencies: one run for every client account, a best-times line per client for the monthly report.
• Creators: scrape your own last 100 posts, see the 3 slots where you do best and the slot where you post most but do worst.
• Brands studying rivals: scrape 10 competitors, get a free all-accounts row with the best times for the whole niche.
• AI agents and n8n or Make flows: run a scraper, then ask "when should we post?" in one call.

How it works (plain words)

1. Each post is compared with the same account's normal (the median of its posts). A post at 2.0x did twice the account's usual engagement. That is why a small and a big account can be pooled fairly.
2. Each post goes into a weekday and hour block (3 hours by default) in your time zone.
3. Each slot gets a score: the median of its posts, pulled towards normal by three imaginary average posts, so a slot with two lucky posts cannot beat a slot with eight good ones. Only slots with 3+ posts are ranked.
4. Days on their own and hour blocks on their own are ranked the same way, because they hold more posts and are steadier than single slots.

Pinned posts (old hits kept on top of the profile) are left out by default.

What you get for each account

Field	What it means
bestSlot, bestSlotLift	The best weekday and hour block, and how many times the account's normal its posts got there
bestSlots	The top 3 slots with posts, median lift and score
bestDay, bestDays	Days ranked on their own
bestHourBlock, bestHours	Hour blocks ranked on their own (any day)
mostUsedSlot, mostUsedSlotLift	When the account posts most now, and how that slot does: the "move from here" slot
worstSlot, worstSlotLift	The weakest ranked slot
confidence	high, medium or low, from posts with a time and posts in the best slot
reason	One plain paragraph with all of the above
heatmap, heatmapPosts	7 days x hour blocks: median lift where a cell has enough posts, and post counts per cell
postsAnalysed, postsWithTime, postsPerWeek, baselineMetric, followers	The sample behind the answer

When 2 or more accounts are analysed, one free all-accounts row (scope: "all_accounts") pools them all.

Example output (duolingo's last 50 TikToks, New York time)

{
  "account": "duolingo",
  "platform": "tiktok",
  "timeZone": "America/New_York",
  "postsAnalysed": 49,
  "postsPerWeek": 6.73,
  "bestSlot": "Fri 09:00-12:00",
  "bestSlotLift": 4.32,
  "bestDay": "Thu",
  "bestHourBlock": "15:00-18:00",
  "bestHourLift": 1.64,
  "mostUsedSlot": "Thu 12:00-15:00",
  "mostUsedSlotLift": 1.39,
  "worstSlot": "Thu 09:00-12:00",
  "worstSlotLift": 0.26,
  "confidence": "high",
  "reason": "Best slot: Fri 09:00-12:00 (America/New_York). Posts there got 4.32x the normal engagement (4 posts). Best day Thu (1.26x). Best hours 15:00-18:00 on any day (1.64x). Most posts go out Thu 12:00-15:00 (6 posts, 1.39x). Weakest: Thu 09:00-12:00 (0.26x). Confidence high.",
  "bestSlots": [
    { "slot": "Fri 09:00-12:00", "posts": 4, "medianLift": 4.32, "score": 2.31 },
    { "slot": "Thu 15:00-18:00", "posts": 4, "medianLift": 2.04, "score": 1.5 }
  ]
}

Which scrapers it reads

Point Datasets at the run of any of these (or a CSV or Sheet export of them). Column names are matched automatically.

• Instagram: apify/instagram-scraper (posts), apify/instagram-reel-scraper, apify/instagram-profile-scraper (latest posts) and look-alikes
• TikTok: clockworks/tiktok-scraper and look-alikes
• YouTube: streamers/youtube-scraper and look-alikes
• X (Twitter): apidojo/tweet-scraper and look-alikes
• Facebook pages: apify/facebook-posts-scraper and look-alikes
• Your own sheet: an account (or username, handle, channel) column, a date column with the clock time, and likes, comments, shares, saves or views. A bare date (no time) counts for the weekday only.

Run it straight after your scraper: on the scraper's Integrations tab, add this actor; the Datasets field is filled with {{resource.defaultDatasetId}} for you.

Input

Field	Default	What it does
Datasets / File or Google Sheet URL / Post rows		Where the posts come from (any mix)
Time zone	UTC	The time zone the answer is given in, e.g. America/New_York, Europe/London, Asia/Tokyo
Hour block size	3	1, 2, 3, 4 or 6 hours per slot. Use 4 or 6 with under 50 posts
Rank posts by	engagement	Engagement (likes + comments + shares + saves), views, likes, comments or engagement rate
Posts needed per slot	3	Slots with fewer posts are not ranked
Posts needed per account	10	Accounts with fewer are written with the reason and not charged
Only posts from the last (days)	0 (all)	Leave out old posts
Leave pinned posts out	on	Keeps the account's normal honest
Add an all-accounts row	on	The free pooled row
Also save as a file		CSV or Excel copy

Pricing

Pay per event, no subscription:

• $0.02 per account analysed (only accounts that come back with a result; too few posts, no dates or a charge limit are written with the reason and not charged)
• $0.01 per export file (CSV or Excel), only if you ask for one
• A tiny Apify start fee per run

Examples: 1 account = about $0.02. A 20-client agency running it monthly = $0.40 a month. 10 rivals plus the free all-accounts row = $0.20. The scraper you run first is billed by that scraper.

FAQ

How many posts do I need? 50 to 100 per account gives medium or high confidence. With 20 to 30, set the hour block to 4 or 6 hours and lean on the best day and best hours.

Is this the time my followers are online? No tool outside the app can see that. This shows when the account's posts have done best, which is the closest public signal, and it is a strong hint, not a guarantee. Big accounts that always post at the same time will show few slots: the "most used slot" line tells you that.

Which time zone should I use? Your audience's. A London brand with mostly US followers should use America/New_York.

Can I check a rival's account? Yes. Scrape their public posts with any scraper above and give this actor the rows.

Does it work for AI agents? Yes. It is pay per event, agent-ready through Apify's MCP server and x402, and takes rows inline in data.

If this told you when to post, a review on the Apify Store helps a lot.

Built by Nero Labs. Questions or a scraper whose columns are not read? Open an issue and it gets fixed.