3 days free on every plan · Try any plan before you pay.

Start free trialStart

BlogCreative testing

Creative fatigue on organic TikTok: do later versions get fewer views?

Does a video wear out as more versions of it go up? Data on 134 videos posted 40+ times per app, next to what TikTok and Meta say about ad fatigue.

By Abondy ResearchPublished 10 min read

In short: In our data, later versions of a video got somewhat fewer views on YouTube Shorts and Facebook, but not clearly on TikTok. We looked at 261 sets of 40 or more posts, each set being one uploaded video's versions on one app. After adjusting for the day each post went up, posts 31 to 40 got about 22% fewer views than posts 1 to 10; adjusting for the time of day as well, about 10% fewer, so the size is uncertain. On TikTok, posts 31 to 40 did about as well as posts 1 to 10 (1.06 vs 1.04 times the video's typical post). These sets went up over a median of 2.4 days, so this is wear inside a short burst, not over weeks.

Creative fatigue is an ads idea: the same people see the same ad again and again, and results slip. Organic posts work differently, because each version goes up on a different person's account and reaches a different crowd. So we asked a narrower question with our own data: when one video goes out as many versions, do the later ones do worse? For how noisy single posts are, and how many posts a fair test needs, see ad creative testing without an ad budget.

What is creative fatigue on TikTok?

Both big ad platforms describe it the same way. Meta's developer page for its creative fatigue alert says: "Fatigue rises as an audience sees the same creative repeatedly, which usually shows up as a falling click-through rate and a rising cost per action." The alert reports a level of LOW, MEDIUM or HIGH (Meta for Developers, read November 21, 2026).

TikTok builds fatigue checks into its ads tools. Its help page for Smart Creative (last updated November 2025) says the tool "observes their performance for 3-5 days, allocating no more than 10% of the ad group's total budget". It then pauses "initial videos that show signs of fatigue within the first 3-5 days" and puts "new videos on the waiting list to replace them" (TikTok Ads Help, read November 21, 2026). The same page says some advertisers will see a newer version called Automate Creative.

TikTok's own research points at the creative more than the repeat count. A May 5, 2025 TikTok article on a System1 study of 8,000 heavy TikTok users in six countries says "High frequency cannot save poorly performing ads." It also reports that "Ads with Brand Recognition in the first 2 seconds generated 57% more happiness and achieved 19% less attention decay" (TikTok for Business, read November 21, 2026).

All of that is about paid delivery to one audience. None of it measures organic posts.

Do later versions of a video get fewer views?

A little, on average. For every uploaded video with at least 40 counted posts on one app, we lined up its first 40 posts on that app in the order they went up. Each version goes up at most once per app, so each post is a different version, in the order those versions went up. Then we compared each block of 10 with that video's typical post on that app.

Posts of one video on one appPostsViews vs the video's typical post, day-adjustedSame, not adjustedDay-adjusted, read posts onlyMedian views (read posts)Share with 100+ views (read posts)Share with 1,000+ views (read posts)
Posts 1 to 102,6101.12×1.18×1.11×128.552.6%26.0%
Posts 11 to 202,6101.00×1.01×1.03×9449.1%26.3%
Posts 21 to 302,6101.04×1.02×1.03×8548.3%25.4%
Posts 31 to 402,6100.87×0.83×0.85×6244.4%23.2%

"Day-adjusted" means each post was first compared with every counted post on the same app that went up the same day (New York time). That takes out swings that hit the whole app on a given day. "Typical post" is the geometric mean, which stops one big post from setting the level. So 0.87× means posts 31 to 40 got about 13% fewer views than that video's typical post.

We never got a view count back for 10.7% of these posts (1,119 of 10,440). They're stored at 0 views, but their real count is unknown, so the last four columns leave them out. They were spread evenly: 9.9% of posts 1 to 10 and 10.8% of posts 31 to 40.

The early posts did best and the fourth block did worst. In 59.4% of the 261 sets, posts 31 to 40 had a lower day-adjusted average than posts 1 to 10. The median set's posts 31 to 40 got 0.78 times the views of its posts 1 to 10. The best posts held up better than the middle: the share of read posts past 1,000 views only slipped from 26.0% to 23.2%.

Is it the version, the day or the account?

Later posts differ from early ones in more than their place in line, so we ran these checks.

  1. The day. Already taken out in the main column. The pattern looks the same with or without it.
  2. The poster account. Some accounts get more views than others, and different accounts may grab a video early. So we also took out each account's own average over all its counted posts, using accounts with 10 or more posts (5,640 of the posts). The blocks came out at 1.14×, 0.98×, 1.03× and 0.86×. Same shape.
  3. Two halves of the sample. We split the videos into two random halves. In both halves, posts 31 to 40 came out at 0.87×. The early lead was less steady: posts 1 to 10 were 1.21× in one half and 1.03× in the other.
  4. Who got the later posts. On the whole, later posts didn't go to smaller or newer accounts. Where we knew them, the median follower count was 6 in both the first and the fourth block, and account age and earlier posts were about the same. So were the share posted between noon and 8pm New York time and how long after posting we last read the views. Versions also went out in no set order: the order of posts didn't follow the order the versions were made in. Taking out each country's average on that app left the first and fourth blocks at 1.11× and 0.88×.
  5. Posts we never read. Leaving them out gave 1.11× and 0.85×.
  6. The time of day. Adjusting for the 4-hour block of the day as well as the day shrank the gap: 1.04× against 0.94×, about 10% fewer. We can't tell whether that check takes out a time-of-day effect or part of a real dip.

So the dip by the fourth block of 10 held up in every check, but its size didn't: from about 10% to 23% fewer views than posts 1 to 10, depending on what we adjusted for. How big the early lead is, we're less sure of too.

We can't say why. One possibility is TikTok's own description of its feed. Its recommendation page says that when top-ranked videos for a viewer are "too similar to each other - for example, they use the same sound - we replace them with others" (TikTok, read November 21, 2026). Something like that could hold back near-identical posts going out close together. But our data doesn't test it, and on TikTok itself we saw no clear dip.

Does creative fatigue differ by app?

Yes, a lot. These are the same first-40-post blocks, day-adjusted, one app at a time.

AppSetsPosts 1–10Posts 11–20Posts 21–30Posts 31–40Sets where 31–40 trailed 1–10Median set: 31–40 vs 1–10
YouTube Shorts1341.09×1.03×1.13×0.79×59.7%0.67×
TikTok691.04×0.93×0.98×1.06×52.2%0.98×
Instagram Reels381.18×0.91×0.95×0.98×63.2%0.90×
Facebook201.57×1.15×0.86×0.64×75.0%0.42×
  • TikTok showed no clear sign of wear within 40 posts. Posts 31 to 40 trailed posts 1 to 10 in 52.2% of sets, close to a coin toss. The checks above ranged from a 7% dip (read posts only) to a 16% rise (time of day).
  • YouTube Shorts held steady for 30 posts, then dipped.
  • Facebook fell the most, block after block, but rests on only 20 sets, so read it as rough.
  • Instagram Reels had a strong first block and then flattened. Our Instagram counts read low (see below), so treat it as rough too.

Counting only posts we read, posts 1 to 10 and posts 31 to 40 came out at 1.06× and 0.82× on YouTube Shorts, 1.58× and 0.54× on Facebook, 1.12× and 0.98× on Instagram Reels, and 1.08× and 1.00× on TikTok.

Do videos wear out over days or weeks?

We can't tell from this data, because a video's versions went up fast. From first post to last on one app, the median set took 2.4 days. Grouped by time since that set's first post, the day-adjusted level was 1.00× in the first 24 hours (8,863 posts) and 1.02× over the next two days (5,264 posts). It was 0.82× from 3 to 7 days in, but that's just 463 posts in 83 sets. Only one post went up a week or more after its set began. So this is about a burst of posts over a few days. It says nothing about running one creative for a month.

How many versions can one video take before it wears out?

From our data, this is what we'd plan for:

  • On TikTok, 40 posts of one video, as different-looking versions on different accounts, showed no clear wear. Spend your effort on getting enough posts per video, not on swapping too early.
  • On YouTube Shorts and Facebook, expect the fourth block of 10 to come in lower. If a video has had about 30 posts on one of these apps within a few days, bring in a new video.
  • Swap the idea, not just the look. A new video means a new hook, angle or format. Versions of one video change the crop, layout, caption and sound; what makes a version count as new is in AI video variations.
  • Judge by medians in blocks of 10, compared with other videos posted the same days. A single post, or the best post, can't show wear. Most views on a short post arrive in the first day, as when views stop coming in shows, so a block can be judged a day or two after its last post, a little longer on Reels.

Where Abondy fits

Abondy is built for this kind of volume. Our video model Flagship 1.2 turns one upload into many versions, each with its own crop, layout, caption and sound, and real people post them on their own accounts. When a video wears out, you upload the next one. Each plan includes free videos to make versions from: 100 a month on Plus, 1,000 on Pro and 2,500 on Ultra. A post costs 12 credits, about 16¢ to 20¢ a post on a plan, and since November 20, 2026 you're charged only for posts that reach 100 views within 24 hours of the link being handed in. Plans start at $20 a month. We make Abondy, so weigh our part with that in mind.

How we counted

  • Sample: counted posts made from August 29 to November 12, 2026 (New York time), pulled on November 21, 2026. Every post was at least 7 days old when pulled.
  • Sets: an uploaded video's versions on one app, with 40 or more counted posts on that app. That gave 261 sets from 134 uploaded videos, 14,591 posts in all; the tables use each set's first 40 posts (10,440 posts). The sets came from 6 campaigns, so they aren't independent: one brand's campaign accounts for about half the posts in the tables (52%), and Abondy's own two campaigns for about a quarter (24%). Read the rows as rough.
  • Who's in it: mostly gaming clips, largely Roblox, posted on small accounts.
  • Never read: 10.7% of the posts in the tables never gave us a view count and are stored at 0. The medians and shares use only the posts we read; the ratios are shown both ways.
  • Views: each post's count when we last read it, not a lifetime total. A few posts were last read before they were 7 days old.
  • Exclusions: Abondy's own test accounts, posts whose views were confirmed as bought, and posts that were flagged, rejected or deleted.
  • Instagram: read the Reels numbers as a floor. Some counts come from each post's public page, and in a 16 Nov 2026 check of 37 posts that page showed about a third of the real count.
  • Associations, not proof: these numbers describe our posts. They don't show why later posts did worse, and they don't promise what your videos will do.

Questions people ask

What is TikTok ad fatigue?

It's when an ad's results slip because the same people have seen it too often. Meta describes it as a falling click-through rate and a rising cost per action. TikTok's Smart Creative watches new ad videos for 3 to 5 days and pauses ones that show signs of fatigue.

Do organic TikTok videos get creative fatigue?

Not clearly, within 40 posts in our data. When one video went out as 40 or more versions on TikTok, posts 31 to 40 did about as well as posts 1 to 10. On YouTube Shorts and Facebook, posts 31 to 40 came in lower.

How often should I refresh creatives for organic posts?

Going by our data, refresh the idea after about 30 posts per app on YouTube Shorts and Facebook. On TikTok, we saw no clear sign of wear within 40 posts of one video.

Does posting more versions of the same video hurt it?

Not in a way we could see on TikTok. On other apps the later posts did a little worse on average. Across all four apps, the share of read posts past 1,000 views barely moved (26.0% in posts 1 to 10, 23.2% in posts 31 to 40).

Read next