The short answer
New UGC accounts go viral when the format changes, not when the account grows.
Read the full answer
Two study-app accounts that started from zero in September 2026 spent weeks at 500 to 1,000 views a reel, then broke out with a 26 to 36 second video that named its audience in the first frame, was built to be sent to a friend and made the app the payoff.
One reached 56.8M views and 1.4M shares, the other 11.1M, and both creators re-ran the winning format within days.
Neither account grew into its big video. Both sat near zero, then one format change moved views by four to five orders of magnitude in a single post.
Source: Instagram public reel counts for @ivy_learns and @studywith.karen, read 2026-11-14 · log scaleTwo Instagram accounts in the AI study-app niche started from zero in September 2026. Within two months one had a reel at 56.8M views and the other a reel at 11.1M. Neither account had an audience when it started, and neither grew into its biggest video. Both spent weeks posting daily at a few hundred views, then changed the format.
This is a reel-by-reel breakdown of what changed, built from the public counts on both accounts as of November 14, 2026. It sits alongside our UGC playbook, which covers the same niche from the outlier side. Only videos past a million views count as evidence here.
What happened on @ivy_learns and @studywith.karen?
@ivy_learns and @studywith.karen both posted one or more study-app videos a day from mid-September 2026, and both broke out after two to four weeks of near-zero views. Each account features one app: Astra AI on @ivy_learns, and Solvely on @studywith.karen.
| @ivy_learns | @studywith.karen | |
|---|---|---|
| First reel | Sept 13, 615 views | Sept 17, 539 views |
| Median of first 12 reels | 665 | 734 |
| First breakout | Sept 29, 493K | Oct 15, 11.1M |
| Biggest reel | Oct 13, 56.8M | Oct 15, 11.1M |
| Reels posted to Nov 13 | 62 | 71 |
| Reels past 100K | 11 | 9 |
| Followers on Nov 14 | 39.1K | 16.4K |
The two paths differ in shape. @ivy_learns stayed flat at a few hundred views for almost two weeks, then jumped. @studywith.karen climbed slowly, from about 600 views to about 13K, before one video went to 11.1M.
The floor moved too. The median of @ivy_learns’ last 14 reels is 11,050 views, against 665 for her first 12. On @studywith.karen it went from 734 to 5,584. A big hit leaves the account with a bigger audience for the next video, but it never produced a second 50M video on its own. The later hits came from running the winning format again.
What changed between the 500-view videos and the viral ones?
The videos got shorter, carried one idea instead of a story, and named who should watch in the first frame. That was the whole change, and both accounts made the same one.
@ivy_learns’ first reel is 101 seconds of storytime about a classmate, with the app nowhere near the opening. It got 615 views.
Her first breakout, 16 days later, is 31 seconds: she holds up a printed anatomy chart under the text “Stop wasting money on printed anatomy cards”, then “Do this instead”. It got 493K.
Two weeks after that she posted 26 seconds that open on her face, eyes wide, holding an iPad toward the camera under the text “All medical students, ATTENTION”. That one reached 56.8M.
@studywith.karen’s first reel is a 43-second list of ADHD symptoms. Her breakout is 36 seconds titled “studying like I’m listening to JUICY gossip”, in which she reads her biology notes in a gossip voice with a pen in her mouth.
Three things hold across all four winners:
- They run 25 to 36 seconds. The flops were 43 and 101.
- The first frame is a face doing something. Wide eyes, a hand on the head, a pen in the mouth. A neutral talking head opens neither breakout.
- The text names a person or a feeling the viewer recognizes, like a medical student or someone bored by their notes, before it says anything about studying.
Why did these videos get shared so much?
The viral videos were made to be sent to one specific friend. “All medical students, ATTENTION” is an instruction to forward the video to the medical student you know, and the 56.8M reel collected 1.4M shares against 2.4M likes. The 11.1M gossip video had 388K shares against 633K likes.
Instagram counts a share each time someone sends a reel on. A reel that keeps getting sent reaches people who never followed the account, which is how a month-old account ends up in front of tens of millions. The 493K breakout shows the same pattern at a smaller size: 13.5K shares against 13.8K likes, nearly one send per like.
A brand reading only likes or views would have seen these as good videos. Reading shares, they were obviously different. If you track one number on creator videos in the first hours, track sends per view, because a video that people forward is still finding new viewers when every other number has stopped moving.
The callout works because it narrows the audience. “All students” would have been ignored. “Medical students” and “Biology students, you NEED this” (another @ivy_learns reel) pick a group small enough that a viewer thinks of a real person to send it to.
Where was the product in the viral videos?
The product sits in the second half, as the payoff. Both 10M-plus videos are product videos, and neither leads with it.
The 56.8M reel spends its first seconds on the creator’s face and the callout, with the iPad held low in the frame.
Then the iPad fills the screen: a 3D human body from the app, which she turns and zooms with a finger down to the heart, under the text “Human Body Atlas, App name: Astra ai”. The app is what the viewer stays to see.
The 11.1M reel opens on the gossip bit and ends on a laptop screen showing her notes turned into a labelled cell diagram, with the caption “thanks solvely ai lol”.
That order matters more than the production. Most product briefs put the product in the first three seconds and the benefit after it. These videos put the viewer in the first three seconds, promise something, and let the product deliver it on screen. Nobody in either video says how good the app is. They show it doing one thing.
It also explains why these formats suit an app with a visible moment. A 3D model you can turn with a finger is worth watching on a phone. A product whose result takes weeks to show would need a different payoff.
What did both creators do after the first hit?
Both creators ran the format again within days. @studywith.karen posted the gossip format the day after her 11.1M video and got 1.8M, then again the day after that and got 882K. Five of her next eight reels passed 200K.
@ivy_learns remade her 56.8M video six days later in Tagalog, “Bakit ngayon ko lang nahanap ang app na ’to?” (“Why did I only find this app now?”), with the same shocked face and the same iPad. It reached 10.8M. Two and a half weeks later the English version, “Why did I only find this app now?”, got 1.2M.
The lesson for anyone paying for creator videos is that a winning format is the cheapest brief there is. It has already told you what the audience responds to, and the second and third versions still beat the account’s typical reel by 50 times or more.
The Tagalog remake also shows the format travelling across languages: the creator is in Manila, and a local-language version reached an audience the English one had not.
Formats do wear out. Both accounts went back to thousands of views between hits, and the remakes earned less each time. Run the winner again while it is fresh, then change one element at a time: the language, the group you call out, the setting.
What does this mean for a brand running creators?
For a brand running creators, it means paying for formats and volume, not follower counts. The app behind @ivy_learns got 71.6M views in 62 days from one creator posting daily on an account that started at zero. Almost all of it came from three videos in one format.
A brand can set up the same conditions on purpose. Brief creators on a format, not a script, and leave them room to say it in their own words.
Ask for one video a day for at least four weeks, because both breakouts here came after two to four weeks of flat numbers that would have ended most test campaigns. Track shares per view on every video in its first day.
When one breaks out, brief the same format to every other creator on the program that week, including local-language versions where you sell.
None of this needs famous creators. Both accounts here were new, both were in a niche that at least eight apps were recruiting for, and the results came from the format and the repetition. Our UGC playbook shows how to find the format in your own niche before you brief anyone.
Impact Origins runs creator programs this way for consumer brands and apps spending $10K or more a month on paid social. We have worked with 50+ brands and delivered 210M+ views across creator programs. Tell us what you’re running.
Where does this breakdown stop?
The breakdown stops at views. Nothing here measures installs, signups or revenue for Astra AI or Solvely, and a 56.8M-view video can reach many people who will never use a study app. We also do not know whether either brand put paid spend behind any of these videos.
Two accounts are a small sample. The pattern is consistent across both and with the outliers in our playbook, but it is a pattern from one niche, studied in one season.
Some details may matter and cannot be tested from the outside. @ivy_learns moved her posting time from early afternoon in Manila to US morning hours on September 26, three days before her first breakout. Two accounts cannot tell you whether that played a part.
We also see only what was posted, not how many takes or rejected drafts sat behind each video. All counts are as Instagram showed them on November 14, 2026, and they will keep rising.
Frequently asked
How long does it take a new UGC account to go viral?
In these two cases, two to four weeks of daily posting. @ivy_learns posted every day from September 13, 2026 and hit 493K views 16 days in, then 56.8M a month in. @studywith.karen posted daily from September 17 and hit 11.1M four weeks in. Neither breakout came from the account growing first.
Both came from a new format.
Do UGC videos need to hide the product to go viral?
No. Both of the 10M-plus videos in this breakdown show the app.
They open on the viewer, not the product, and put the app in the second half as the payoff: a 3D heart you can turn with a finger in one, and notes turned into a diagram in the other. The product is the reason to keep watching, not the opening line.
What should a brand do when a creator video goes viral?
Run the same format again within days, on the same account and on others. @studywith.karen posted her 11.1M format again the next day (1.8M) and the day after (882K). @ivy_learns remade her 56.8M video in Tagalog (10.8M) and later in English (1.2M).
A format that has just won is the cheapest next video a brand can brief.
Which metric predicts whether a UGC video will spread?
Shares. The 56.8M reel had 1.4M shares against 2.4M likes, and the 11.1M reel had 388K shares against 633K likes.
Instagram counts a share each time someone sends the reel on, and a video people forward keeps reaching new viewers after its own followers have seen it. Likes alone would not have flagged either video early.



Affiliate Marketing vs UGC: Which Should a Consumer Brand Spend On?
What Are the Best Influencer Marketing Agencies for Consumer Brands in 2026?
What Are the Best UGC Ad Agencies for Consumer Brands in 2026?