How we estimate a trend's views

How we estimate a trend's views

Each trend page carries one line like this, here for Bulin 47: About 370M views across TikTok, Reels, Shorts (likely 210M–745M), as of 16 Oct 2026. Every piece of that line means something exact:

• About 370M is our middle estimate for every copy of the trend added together, with the viral post they copy when it's a social post. It isn't the play count of any one post.
• Likely 210M–745M is the range we think the real total falls in. The low end sits close to what we could read post by post; the high end allows for what we couldn't see.
• TikTok, Reels, Shorts names the apps the estimate covers. Reels are estimated rather than read (see below). I Wanna Love You names only TikTok, because we haven't estimated its Reels.
• As of 16 Oct 2026 is the day of our last read. Anything we couldn't re-read that day keeps its own, older date in the trend's file.

What we count, and what we leave out

We count posts that are the trend itself. We leave out anything that only shares its song or its hashtag.

The long tail is the part no one can list: thousands of copies too small to turn up anywhere. Post sizes fall off in a regular way from the biggest down, so we fit a curve to the posts we did read and run it out to the number of posts the sounds and tags show, or to a range when they show none. For Rumpelstiltskin, assuming anything from 2,000 to 10,000 posts barely moved the result, because the smallest ones add so little.

• Counted: the post the trend copies. Bulin 47's street freestyle had 8.2M plays on TikTok on 16 Oct, and the same clip went up on Instagram.
• Counted: the copies, one at a time. For Rumpelstiltskin we read 53 public TikTok posts, 35.05M plays between them on 16 Oct.
• Counted: the trend's own sounds and tags. About 18,000 videos used Bulin 47's five trend sounds on 14 Oct. Counts like that tell us how far the curve has to run.
• Counted: Shorts, and X when a big copy turns up there. Rumpelstiltskin's total includes one post on X.
• Left out: the music video, film or show behind a trend. Raindance's music video had more than 285M views on YouTube, Hotel Lobby's COLORS performance more than 131M and the Burning Bridges video about 27M. None of that is in their totals.
• Left out: lip-syncs and other memes on the same sound or tag. Most of the 6,476 videos on I Wanna Love You's three sounds (15 Oct) are people lip-syncing, not the cat. Burning Bridges leaves out a separate real-people meme of about 90M views.
• Left out: reposts, and real people as themselves. Alors on Danse doesn't count reposts of the 2010 sketch it copies.

The eight live trends' estimates, read on 16 Oct 2026

Biggest first. The estimate and range are the ones on each trend page. How sure is the label we write with each estimate, and the last column is the biggest thing it can't see or chooses to leave out.

Each trend's file, under Sources, lists every post we read with its count and date. No app publishes a trend's total across every copy, so we can't yet check these against a true figure.

Why the top posts alone undercount a trend

Our first version, read on 11–12 Oct 2026, fitted a curve to the 5 or 6 biggest posts we could see and assumed how many posts there were. On 14 Oct we rebuilt it from the trend's sounds and tags, many more copies, YouTube Shorts and the original posts. The two biggest trends came out about 18 and 19 times higher.

The first version missed Bulin 47's original on Instagram, the thousands of videos on the trends' sounds and tags, and every copy beyond the handful we started from. Now a sum of the posts we read is a floor, not an estimate: for Rumpelstiltskin and I Wanna Love You, the low end of the range is about that sum.

Are TikTok view counts accurate?

Accurate enough to add up, but rounded, and they miss what no one can see. Once a post passes a million, the public count moves in steps of 100,000. Bulin 47's original read 7.9M on 15 Oct and 8.2M on 16 Oct, so it grew by anywhere from about 2.5% to 5%. We take the low end of a rounded change: we raised its estimate by 2.5% and rounded down, to 370M.

Posts also drop out between reads. Alors on Danse's rappers version had 305K plays on 15 Oct and didn't show on its creator's profile when we re-read it on 16 Oct. A post we can't re-read keeps its last count and that day's date; we never copy it forward as a new reading. Private posts never show at all, so every total misses some.

How are TikTok and YouTube Shorts views counted?

Each app decides what a view is, and we take the number each post shows. YouTube changed Shorts on 31 Mar 2025: a view now counts each time a Short starts to play or replays, not after a set number of seconds. TechCrunch, reporting the change, said TikTok and Instagram already counted that way. The older Shorts figure lives on as engaged views in YouTube Analytics.

So, by TechCrunch's account, a replay adds a view on all three apps, and our totals add the apps' numbers as they stand. Shorts copies are a small share of most trends here. When we re-read Hotel Lobby's top 20 Shorts on 15 Oct they came to 0.61M, not the 4.0M we had on 14 Oct, so its estimate went from 140M down to 137M. Estimates can fall as well as rise.

Why we can't see views on Instagram Reels

Logged out, which is how we read every post, Instagram shows a reel's likes but not its plays. That makes Reels the least certain part of each estimate. We estimate them, from the likes or as a share of the TikTok total, note the assumption in the trend's file, and widen the range to cover it.

It's also a big part of why some ranges are wider than others. Bulin 47's high end is about 3.5 times its low end, partly because the freestyle's Instagram reel, likely its biggest single post, hides its views. I Wanna Love You's is about 1.4 times: nearly all of it is TikTok posts we read one by one, and its low end, 49M, is about their sum.

How to see hashtag views on TikTok, and why a tag isn't a trend

A tag's page on TikTok can show how many videos use it and their total views. We read those pages for every trend, because they're our best guide to how many small copies exist. They don't always answer: on 14 Oct 2026 #bulin47freestyle showed 88.8M views, and on 15 and 16 Oct the tag and sound pages we read showed no counts, so those figures keep their 14 Oct date.

A tag total still isn't a trend total. Some posts on a trend's tag aren't the trend, and plenty of copies skip the tag. So we use only tags made for the trend itself, never a general one like the song's name, and we use them to size the long tail rather than as the answer.

How often we re-read the numbers

Every live trend is re-read on our daily trends run. All eight were read on 15 and 16 Oct 2026, and six of them on 14 Oct as well. Our rule is that no number on a trend page is more than 3 days old. The date in each line is the last read; a page's Updated date moves only when we change its words, not when a number refreshes.

A big total doesn't mean a trend is still moving. Hotel Lobby's copies add up to about 137M views, yet Google searches for it peaked on 26 Sep 2026 and stayed under a tenth of that peak from 11 Oct, and its three biggest copies have shown the same counts since 14 Oct. For how quickly searches fade, see how long TikTok trends last.

Three trends, three kinds of estimate

Each shows a different side of the method. All three start with a free preview: add a selfie, or your pet's photo, and see a still of the result before you pay anything.

1. Bulin 47↗ Rising ≈380M views · 1 person · 15s The biggest estimate and the widest range. About 370M views, likely 210M–745M, partly because the freestyle it copies hides its views on Instagram. It's just you: the red lights, the speaker and the crowd filming on their phones, with you in the middle. See yourself in it
2. I Wanna Love You↗ Rising ≈55M views · 1 pet · 10s The tightest range, 49M–67M. Almost all of it is TikTok posts we read one by one. Your cat or dog takes the singing cat's place as the whole animal, in the same corduroy suit, filming a night selfie. See yourself in it
3. Raindance ≈200M views · 2 people · 28s A clear case of what we leave out. Its 200M estimate doesn't include the music video it copies, though its biggest AI copy alone had 50.5M plays on 16 Oct. It's for two: you and your partner or best friend on the pier at golden hour. See yourself in it

Where these numbers come from, and what they aren't

Every number here comes from public posts, read the way any logged-out visitor sees them. They are our estimates, not figures from TikTok, Instagram or YouTube, and we aren't affiliated with any of them. None of it comes from the people who use this site: what we record about our own visitors is on the privacy page.

Trend pages play a short silent loop of the seconds people copy, and the sound button opens the original post on its own platform; we never re-host a clip with its audio. If you own a video we show and want it removed, the steps are on the takedown page.