The algorithm is simpler than people think
Every platform rewards roughly the same thing: content that holds attention and makes people respond. That's it. If people watch to the end, rewatch, comment, and share — you get pushed to more people. So your entire job breaks into three skills: write hooks that stop the scroll, structure videos that hold attention, and read your analytics to do more of what works. Master these three and growth is just a matter of reps.
Skill 1: Hooks (the most important sentence you'll write)
The first 1-2 seconds decide whether a video lives or dies. If they scroll, nothing else mattered. A hook does one of these jobs:
- Call out the exact viewer: "If you're skinny-fat and stuck, watch this."
- Make a bold/specific claim: "I fixed my jawline in 90 days without surgery — here's the truth."
- Open a curiosity gap: "The one thing that finally got me lean wasn't training."
- Lead with the visual proof: show the before first, talk second.
- Promise a fast payoff: "Three skincare steps that cleared my face in a month."
Rules: be specific (numbers and timeframes beat vague claims), never bury the lead, and never start with "hey guys." Write 5 hook options per video and pick the strongest — hooks are where you should spend the most thinking time.
Skill 2: Retention (keep them watching)
- Front-load value. Deliver the goods in the first sentence; tease the rest. People give you about 3 seconds to prove the video is worth it.
- Cut every dead moment. Tighten edits, kill pauses, no rambling intros. Pace is retention.
- Open a loop, close it at the end. "And the last one is the one nobody does — stick around." Gives them a reason to finish.
- Make the end satisfying. A strong payoff drives rewatches, and rewatch rate is gold to the algorithm.
Skill 3: Analytics (do more of what works)
Don't guess — read the data. Check these weekly:
| Metric | What it tells you | What to do |
|---|---|---|
| Watch time / avg % watched | Was your hook + pacing good? | Low early drop-off = fix your hook. Low overall = tighten edits |
| Saves & shares | Did it provide real value? | High saves = make more of that topic; it's your winner |
| Comments | Did it spark response? | Reply to all of them early; replies boost reach |
| Follows-per-view | Did it convert watchers to fans? | High = make more of that exact content |
| Reach vs followers | Are you getting found by new people? | Mostly-followers reach = your hooks aren't traveling |
Find your top 3 videos every month and ask: what did they have in common? Then make more of that. Most accounts grow by ruthlessly doubling down on their few winners, not by constant novelty.
The consistency multiplier
None of this works without volume. The algorithm needs reps to learn what your audience wants, and you need reps to get good. The accounts that win post consistently for long enough to find their winners. Consistency isn't a virtue here — it's the mechanism.
Field assignment
For your next video, write 5 different hooks and pick the best. After it's been up 48 hours, open your analytics and find your single best-performing post so far. Identify why it worked, and plan your next video to repeat that.