# How much of TikTok is luck? The same video, posted twice

1,102 times a creator posted the exact same TikTok twice: how far apart the views were, and whether reposting helps. Published 2026-11-10 by Yap & Done. Videos posted January 2025 to October 2026.

Canonical page: https://www.yapanddone.com/studies/same-video-twice/ · Data: https://www.yapanddone.com/studies/same-video-twice/data.json · https://www.yapanddone.com/studies/same-video-twice/data.csv · License: CC BY 4.0

## The top lessons

- Don't read too much into one video's views. Post the exact same video twice and the typical gap is 2×; two different videos from the same creator differ 2.6×. Most of the difference is luck.
- Reposting a flop usually beats the flop, but don't expect a hit. When the original got under half the creator's usual views, the repost did better 72% of the time, yet still got -56% views against the creator's usual.
- Don't repost a video that already did well. Reposts of hits beat the original only 16% of the time.
- On average, a repost gets fewer views than the original (-15% at the median), so make something new when you can.

## Key findings

- **The same video got 2× different views.** Across 1,102 pairs of identical TikToks, the typical pair's views differed 2× (range 1.7×–2.3×). Two different videos from the same creator, posted the same distance apart, differed 2.6×. 50% of identical pairs differed 2× or more, 22% differed 5× or more, 10% 10× or more, and 1 in 100 differed 98× or more.
- **Most of the gap between two videos is luck.** Measured on a log scale, the gap between two posts of the same video was about 70% of the gap between two different videos (range 61–83%). The video itself explains the rest, which is why changing the video still matters, but any single result is mostly chance.
- **Reposts usually get fewer views than the original.** The second post beat the first 41% of the time (range 34–49%), and the median repost got -15% views against the original.
- **Reposting a flop beats the flop, but not the usual.** When the original got under half the creator's usual views, the repost beat it 72% of the time but still got -56% against the creator's usual. When the original was about usual, the repost beat it 45% of the time. When the original did well (over twice the usual), only 16%.
- **Same-day reposts are mostly rescues.** Reposts on the same day beat the original 63% of the time, likely because creators repost straight away when the first one stalls. Reposts 1–30 days later beat it 27% of the time, and over 30 days later 37%.

## How far apart identical videos land

Share of pairs whose views differed by at least each amount: the same video posted twice, against two different videos from the same creator.

### Pairs that differed by…

|  | Same video twice | Two different videos |
| --- | --- | --- |
| 2× or more | 50% | 60% |
| 5× or more | 22% | 31% |
| 10× or more | 10% | 20% |
| Typical gap | 2× | 2.6× |

Different videos: the original against the creator's other video posted closest to the repost.

## Does reposting help?

The repost against the original, by how the original did against the creator's usual views.

### Repost beat the original

- Original flopped (under half the usual): 72%
- Original about usual: 45%
- Original did well (over twice the usual): 16%

### Against the creator's usual

| Original | Original | Repost | Pairs |
| --- | --- | --- | --- |
| Original flopped (under half the usual) | -67% | -56% | 174 |
| Original about usual | -1% | -15% | 606 |
| Original did well (over twice the usual) | +495% | +48% | 321 |

### By time between posts

| Reposted | Typical gap | Repost beat original | Pairs |
| --- | --- | --- | --- |
| Same day | 1.5× | 63% | 273 |
| 1–30 days later | 1.7× | 27% | 304 |
| Over 30 days later | 2.5× | 37% | 525 |

## Short-video script

- Hook: Same video, same account, posted twice. One got 10 times the views.
- The number: We found 1,102 TikToks posted twice. The typical pair's views differed 2×, and 1 in 10 differed 10 times or more.
- Show: The "Pairs that differed by" table: same video against two different videos.
- Takeaway: One video's views are mostly luck. Judge your content over 10 videos, not one.
- Credit: Source: Yap & Done study, yapanddone.com/studies/same-video-twice

## FAQ

- **Does reposting a TikTok help?** Usually not: the repost beat the original only 41% of the time. If the original flopped, the repost beat it 72% of the time, but still got fewer views than the creator usually gets.
- **Is going viral on TikTok luck?** Partly. The same video posted twice by the same account differed 2× in views on a typical pair, and 1 in 10 pairs differed 10× or more. That's about 70% of the gap between two different videos.
- **Should I delete a TikTok that flopped?** We can't see deleted videos, but reposting the same video rarely turned a flop into a hit. A new video with a different hook is a better bet.
- **Why did the same video get different views?** Who TikTok shows it to first, what else is posted at the same time and how those first viewers react all vary, so identical videos land differently.

## Method

- Pairs: videos from creators in our 10 niches with the same written description (at least 40 characters, not counting hashtags and tags) and the same length, posted within the same account, where no other creator used that description. Groups of more than two were left out.
- Checking they were the same: we fetched every video again and compared its cover, first frame and animated cover as small greyscale images. A pair counted if it was closer than 95% of pairs of different videos from the same creator. 1,102 of 1,347 pairs passed (82%).
- Views are as fetched in November 2026. Videos under 7 days old were left out, and each video's views were compared with the median of the creator's other videos within 30 days.
- Different videos: for each pair, the creator's other video posted closest to the repost (within 7 days), against the original.
- Ranges are 95% ranges from resampling creators.

## Limitations

- Creators choose which videos to repost, usually ones they think deserved better, so reposts aren't a random sample.
- Deleted videos can't be seen, so delete-and-repost cases only count where both copies are still up.
- Small edits (a new caption, a trimmed second) can pass the image check.
- The creators came from niche searches. Among ordinary accounts picked at random we found only 16 pairs, too few to test.

Cite as: Yap & Done (2026). How much of TikTok is luck? The same video, posted twice https://www.yapanddone.com/studies/same-video-twice/