The short answer
Million-view videos in a niche share a format, not a production style: a hook about the viewer's life rather than the product, a value moment the viewer can watch happen, and the product placed inside advice instead of in front of it.
Read the full answer
In the AI study-app niche, accounts with 5,500 and 4,600 followers reached 1.3M and 1.2M views this way. A brand gets more of these videos by finding the format first, then paying many creators to run it on their own accounts, each in their own words.
Raw views mostly measure the account. Views against the account's own typical post measure the format, and the format is the part you can copy.
Source: vidIQ outlier search and account baselines, AI study-app niche, August 2026Most brands brief creators on the product: the features, the claims, the logo rules, then a request for a video. The videos that reach a million people in a niche almost never start there. They open on the viewer’s problem, and the product shows up later, often as one item in a piece of advice.
This playbook shows how to find those videos in your own niche and how to get creators making them. It is built on one niche we tore down in August 2026, AI study apps, which is one of the most crowded creator categories on TikTok and Instagram. At least eight apps were recruiting from the same pool of student creators at the time.
Only videos past a million views count as evidence here. A video with tens of thousands of views can be luck, but a format that clears a million on an account with a few thousand followers is telling you something about the format.
What do million-view videos in a niche have in common?
Million-view videos in a niche share a format. In the study-app niche it had three parts: a hook about the viewer’s life, a value moment you can watch happen, and the product placed inside advice. These are the organic videos past a million views in our August 2026 pull.
| Video | Views | Hook or format |
|---|---|---|
| @lockedinwithmon | 2.4M | Eats on camera through the pitch |
| Overwhelm confession | 1.8M | “Mom my phone is SAVING my grades!!” |
| Overwhelm confession | 1.6M | “explain to me how some people have time to do EVERYTHING in college” |
| @studytipslogan | 1.5M | Calls out one major, shows three useful sites, the app is the second |
| @xy.top | 1.3M | Anatomy flashcards made by hand from cardboard, then self-tested |
| @moushicastudiess | 1.2M | “the smartest people i know are lazy asf”, eating on camera |
Not one of those hooks names an app. They are about grades, time and being overwhelmed, so a student recognizes their own week in the first second and stays to see where it goes.
The anatomy video shows the value instead of claiming it. A student made flashcards out of cardboard, by hand, and tested themselves on camera. That is the exact job the AI apps in this niche do automatically, and the video did 1,213 times that account’s usual views.
Studley AI’s house format, the 1.5M video in the table, hides the product in plain sight. It lists three useful websites for a named major and puts its own app second, so the list reads like a tip from a classmate.
On two of the six, the creator is eating on camera. The food has nothing to do with studying. It gives the eye something to follow while the creator talks.
Why does follower count matter less than the format?
Short-form feeds rank each video on how viewers react to it, so a small account with the right format can outrun a large one. The two cleanest examples in the study-app niche came from accounts with 5,500 and 4,600 followers, and they reached 1.3M and 1.2M views.
The same holds for brand-new accounts: two study accounts that started from zero in September 2026 reached 56.8M and 11.1M views within two months, and the reel-by-reel breakdown shows what changed.
That makes raw views the wrong number to sort by. A million views on an account with two million followers tells you about the account.
A million views on an account that usually gets a thousand tells you about the video.
In the teardown, every video is scored against its own account’s typical post: the anatomy flashcard video beat its baseline 1,213 times over, and Studley’s 1.5M video ran on an account whose typical post got one to two thousand views.
It also changes what a brand should buy. If reach comes from the format, the thing worth paying for is the format plus enough creators to run it, rather than one creator with a big audience.
The study-app category worked this out early. When Quizlet announced it had bought Coconote in early 2026, it pointed to more than a billion views of Coconote’s creator-led social content over the previous 18 months.
At least three study apps ran creator accounts of their own. One of them, with the brand’s name in the handle, posted the same two scripts with at least five different women on camera.
Public view counts can’t show whether paid spend sat behind a video, because a boosted post reports paid and organic views as one number. The biggest video in the niche, 14.8M views on “Finals are here and I don’t know where to start”, was a paid partnership with Google Gemini. We keep sponsored outliers in a separate column and never use them to judge a format.
How do you find the million-view format in your own niche?
To find the million-view format in your niche, pull the outliers, score each one against its own account, watch every video, and group them by mechanism. The study-app teardown ran in this order, and the steps carry over to any consumer category where creators post about products.
- Search your category for videos from the last three to four months with high views. We used vidIQ’s outlier search, limited to videos posted after May 1, 2026 and aimed at US college students.
- Divide each video’s views by its account’s median, and sort by that multiple instead of raw views.
- Pull the numbers a second time. Two entries on our first study-app list did not reproduce, so they came out.
- Watch every video and write down the hook word for word, the first three seconds, the moment the value appears, where the product shows up, the runtime, and roughly how long it would take to shoot. Most of the study-app outliers could be shot within an hour with a phone, a screen recording and decent light.
- Group the videos by mechanism. The study-app outliers fell into five hook families: wish-I-knew advice, the shock reveal, contrarian takes on ChatGPT, the overwhelm confession, and the third-party observation (“that one girl in my college class”).
- Check how worn out each format is. One that a hundred accounts have run for a year is spent. One that peaked about six months ago and went quiet is often worth reviving.
What comes out is a short list of formats, each with its evidence, its mechanism and a call on how fresh it is. In the study-app niche the hand-made anatomy cards stood out. They had the highest multiple in the dataset, they show the product’s job without describing it, and no app was running the format.
How do you get creators making those videos?
Brief the format, not the script, then run it across many creators and change one thing at a time. A creator reading a brand’s lines sounds like an ad within a second, while a creator who understands the mechanism and the value moment will say it in their own words.
A format brief covers the hook mechanism (an overwhelm confession, say), the value moment that has to be on screen, the point where the product enters, and what must not happen. The first thing on that last list is naming the product in the opening line.
When a hook has been copied too often, keep the mechanism and change the surface: a different speaker, a different setting or a different subject.
Then run it across as many creators and accounts as the budget allows. Studley AI ran its house format across at least four creators.
Turbo, another study app, asked creators to open new TikTok and Instagram accounts and post about the app most days. Volume matters because single videos swing so much: the 1.3M anatomy video came from an account whose typical post got about a thousand views, and nobody could have picked it in advance.
Once a format hits, test inside it one variable at a time. Try who is on camera, how big the payoff is, the hook text and where it sits on screen, the caption, the posting time. Change two at once and you won’t know which one moved the number.
Check who you are recruiting, too. When we built a creator list for the study-app niche, four of the accounts on it turned out to be run by competing apps, and a message to any of them would have gone straight to a competitor’s marketing team.
Where does this playbook stop working?
The playbook stops at attention. It finds formats that reach people and says nothing about what those people buy. Nothing in this teardown measures signups, sales or return on spend, and a million views should never be read as revenue. A video can reach a huge audience in the wrong country, at the wrong age, or full of people who will never be in the market.
Formats also wear out. Once enough accounts copy a hook, viewers scroll past it on recognition, so expect a few good months from a winning format before it needs changing.
And the product needs a value moment that fits in a few seconds. The study-app formats work because a deck appearing or a quiz getting graded is visible on a phone screen almost instantly.
A product whose value takes weeks to show, like a supplement, needs a different kind of format, usually built around a person and their routine. A good teardown says so instead of forcing a format that doesn’t fit.
What does a niche teardown from Impact Origins include?
A niche teardown is this playbook built for your category. You get the million-view outliers in your niche with links, each scored against its own account, the hook families behind them, a call on which formats are fresh and which are spent, and a format brief for each one worth running.
It is the right first step when you are about to brief creators and don’t know yet what wins in your niche, or when your current creator videos all look like ads. If you already have a format that works and need more people shooting it, skip the teardown. What you need is a campaign.
Campaigns are for consumer brands and apps spending $10K or more a month on paid social. Impact Origins has worked with 50+ brands and delivered 210M+ views across creator programs, and we reply to every inquiry within 24 hours. Tell us what you’re running.
Frequently asked
What is a UGC playbook?
A UGC playbook is a written record of which video formats reach the most people in one niche and how to brief creators to make them.
It lists the outlier videos, the hook and value moment each one uses, which formats are still fresh, and what a creator needs to shoot one in their own words. It has to be built per niche, because the formats that win for a study app are not the ones that win for skincare.
How do you find viral UGC formats in your niche?
Pull the outlier videos in your category from the last three to four months and rank them by views against the account's own typical post, not by raw views. Watch each one and write down the hook word for word, the first three seconds, the moment the value appears and where the product sits.
Then group the videos by mechanism. A format that shows up on several unrelated accounts is the one worth briefing.
Do UGC creators need a big following to get a million views?
No. In the AI study-app niche, an account with 5,500 followers reached 1.3M views and one with 4,600 followers reached 1.2M.
Short-form feeds rank each video on how viewers react to it, so a small account running the right format can outrun a large one. That is why a creator program should pay for formats and volume, not follower counts.
How long does a viral UGC format keep working?
A few months, in our experience. Once enough accounts copy a format, viewers start scrolling past it on recognition.
Keep the mechanism and change the surface: a new authority figure, a new setting, a new subject. Formats that peaked about six months ago and then went quiet are often worth reviving.



Affiliate Marketing vs UGC: Which Should a Consumer Brand Spend On?
Why Aren't My App Installs Converting Into Trials?
How Is ASO Different on the App Store vs Google Play?