Module 3: Study the field before you play

Feed Training

Your feed is a research tool if you train it. Feed Training means deliberately teaching the platform's recommendation system to show you your niche: search your topic, watch the top videos all the way through, follow the best creators, and briskly skip anything irrelevant. Within days, your feed becomes a live focus group of what's working in your lane.

Do this on a dedicated stretch of time — 20 focused minutes beats two hours of half-scrolling. You're looking for patterns: which openings hook you, where you get bored, what makes you watch twice. Notice your own behavior as a viewer; it's the cheapest market research on earth.

After a week of training, your feed should feel like a trade magazine for your lane. If it doesn't — if it's still full of unrelated viral junk — you're skipping too slowly or watching too passively. Be ruthless: the algorithm learns from every second you linger, so linger only on your future.

Train the search side too: use the platform's search bar for your niche terms weekly and watch what autocomplete suggests — those suggestions are literally the phrases your audience types. Make videos that answer the top suggestions and title them with the exact phrasing. You're not guessing what people want; the search box is telling you, in their words, ranked by volume.

Steal the structure, not the video

There's a bright line between studying and copying. Copy the video and you're a knockoff; copy the structure and you're a student. Structure means: how the first 3 seconds open, where the tension comes from, how often the shot changes, where the payoff lands. Those are mechanics, and mechanics are free to learn from.

Pick one video that performed well in your niche and dissect it on paper. Write out the beats: second 0–3 (the hook), the setup, each new piece of information or joke, the ending. You'll find most viral videos in a niche share a skeleton — once you see it, you can hang your own material on it endlessly.

Dissect across niches, not just yours. The best hook structure you ever steal might come from a cooking video when you make fitness content — structure transfers, content doesn't. Once a month, spend 20 minutes on a completely unrelated niche's top videos and ask "what's the skeleton here?" Cross-pollination is where original-feeling formats come from; everyone in your lane is studying the same five creators.

Build your swipe file

A swipe file is your personal library of proven ideas: hooks that stopped your scroll, formats worth adapting, captions that sparked comments. Save videos into named collections (most platforms have a save/bookmark feature) and add a one-line note on why each one worked. The note is the valuable part — without it, you'll revisit a saved video in a month and have no idea what you saw in it.

Organize by job, not by topic: "hooks," "endings," "series ideas," "things my audience argues about." When you're stuck on video 14 and the blank screen is winning, your swipe file is where ideas come from on demand.

Review it weekly for 10 minutes. You'll start seeing the same structures repeat across niches — that's your education compounding. The creators who seem endlessly creative usually just have the best-organized swipe files. And when a saved video's trick stops working, note that too: a swipe file that records what died is as valuable as one that records what worked.

Share swipe files with one trusted creator friend in your lane. You each see different things, and trading notes doubles both libraries. Set a monthly 30-minute call: three best finds each, why they worked, one idea you're stealing the structure from. Iron sharpens iron, and two organized minds beat one — plus the accountability keeps the habit alive past the first enthusiastic month.