How I Predict Viral Videos Before They Explode

$11,900 last month. Four tools. I know a video will hit before I build it.

Right now someone in your niche is posting the video that pulls 2M views this week. It already exists as a pattern. Most people scroll past it. A few spot the shape, copy it onto a fresh subject, and ship before the trend cools.

The window on any format is 2 weeks. Then everyone floods it and it dies. Speed is the edge.

Here's the full build.

Why Most Faceless Channels Die

They pick a topic they like. They make a video they think is good. It gets 400 views. They quit.

The editing was never the problem. The input was.

A viral video is a format that already worked, pointed at a fresh subject. The format is the asset. The subject is disposable.

Find the format that is hitting this week. Feed it a new subject before the window closes.

That's the model.

The Pipeline

find the format → write the script → build the video → publish everywhere → track what hit

Research → Claude → CapCut → Make → ViewsFour tools. The research runs first, every time.

Step 1 - The Research Engine (this is the whole game)

This step is 80% of the result. Skip it and the other three steps produce polished videos nobody watches.

I don't ask "what should I post." I ask "what is already outperforming, and why."

Here is the exact method.

Build a swipe feed

Track 40 faceless channels in one lane (AI stories, history facts, motivation edits, pet clips, whatever you ship). Open a Google Sheet. Every day, log any video that broke out.

Score the outliers

An outlier is a video pulling far more views than the channel's own average. The math:

outlier score = video views ÷ channel median views

A channel that averages 20,000 views drops a video at 900,000. Score of 45. That video found something. Above 10 is a signal. Above 30 is a format you copy this week.

Ignore the raw view count. A 2M-view video on a 5M-average channel taught you nothing. A 900k-view video on a 20k channel is a map.

Separate topic from format

The topic is what the video is about. The format is the shape it's built in. Topics die in a week. Formats repeat for months.

Example. "POV: you're the last human on Earth" hits. The topic is loneliness. The format is "POV: you're the last X in Y." That format reskins into 200 videos: last knight, last barista, last dragon, last dial-up user. One outlier, a season of content.

Read the first 2 seconds

The hook does most of the work. For every outlier, log two things: the exact first frame and the first line of text on screen. That is what stopped the scroll. The rest of the video only has to not disappoint.

Feed the pattern to Claude

Paste 20 outlier titles and hooks into Claude and let it extract the shape, then generate new ones on your subject.

Prompt for pattern extraction:

1You are a short-form video strategist.
2
3Below are 20 videos that outperformed their channel by 10x or more.
4For each one I've listed: the title, the first line of on-screen text,
5and the outlier score.
6
7[paste your 20 rows]
8
9Do three things:
101. Group them into 5 repeatable FORMATS (the shape, not the topic).
112. For each format, write the hook rule in one sentence.
123. Rank the 5 formats by how easily they reskin to a new subject.
13
14No general advice. Only patterns you can see in this exact data.

Prompt for idea generation:

1Take format #1 from your last answer.
2My subject is [your niche: e.g. ancient Rome facts].
3
4Generate 30 video concepts in that exact format.
5For each: the on-screen hook line (max 8 words), and the one
6image or moment that carries the first 2 seconds.
7
8Rank them 1-30 by scroll-stopping power. Cut anything generic.

Now you have 30 concepts that share a proven skeleton. You didn't guess once.

Step 2 - Formats That Keep Hitting

You will see the same shapes surface every month. Log them once, reuse forever.

**POV: you're the [role] when [event] **

Puts the viewer inside the scene in 1 line.

**What if [familiar thing] but [twist] **

Curiosity gap, no context needed.

**The invisible-force gag **

Objects move on their own, a chaotic reveal at the end.

**Oddly satisfying process **

A start-to-finish build or transformation, no talking.

**1 vs escalation **

1 vs 100, level 1 vs level 100, small stake to huge stake.

**The creature reveal **

A normal scene, then the last 2 seconds flip it.

**Day in the life of [unexpected subject] **

A rock, a medieval peasant, a deep-sea fish.

Each of these carried a different subject to millions last month. The shape is the reusable part. Your subject is the variable you swap in.

Step 3 - Claude Writes the Script for Shots, Not Prose

Once a concept is picked, Claude writes to the format, not to a blog post. It writes for the cut.

Prompt for a short script:

1You are writing a 25-second vertical short in the format:
2"POV: you're the last lighthouse keeper on Earth."
3
4Write for shots, not paragraphs. Output:
5- on-screen hook text for the first 2 seconds (max 8 words)
6- 5 shots, one line each, described for AI image generation
7- one line of narration per shot, spoken pace
8- the final 2-second beat that makes people rewatch or comment
9
10Tone: quiet, eerie, no filler. End on an open loop, not a bow.

Claude questions the gaps if the concept is thin, then hands you a shot list you can build directly.

Step 4 - CapCut Builds the Video

CapCut's AI tools turn the shot list into the finished vertical. Generate the frames, animate them, drop in AI voice and captions, export 9:16.

Instruction for the build:

1Build a 25-second vertical (9:16) short from this 5-shot list.
2Generate each shot as a moving clip, 5 seconds each, back to back.
3Add the hook text over shot 1 for the first 2 seconds, large, high contrast.
4Auto-caption the narration, word-by-word highlight, bottom third.
5Score it with a low, tense ambient bed that never covers the voice.
6Hold the final shot on the open loop. Export for TikTok and Shorts.

Lock your recurring characters and settings as assets so a one-off clip becomes a repeatable series. Cast once, change the plot, ship forever.

Step 5 - Make Publishes and Tracks What Hit

The research loop only compounds if you feed your own results back into it. Make closes the loop.

The scenario runs on a schedule:

Trigger: new finished video in a Google Drive folder

1. Read the concept, format tag, and hook line from the intake row
2. Post the vertical to TikTok, YouTube Shorts, and Instagram Reels
3. Write titles and tags from the format tag
4. Wait 48 hours, pull view counts back via each platform's API
5. Write the new outlier score into the same sheet you research from
6. Flag any of your own videos that beat your median by 10x
7. Send a Telegram ping: which format won this week

Now your own hits become inputs. The formats that work for your account rise to the top. The factory learns.

Setup time: 4 hours once. Runs indefinitely after.

How You Make Money

**Platform payouts **

Short-form funds and AdSense pay on views. A faceless channel that lands 3 to 6M monthly views clears $2,000 to $6,000/month. The research engine is what moves you from 20k-view floors to breakout ceilings.

**Faceless channel sales **

A channel with a proven format and 100k followers sells for $3,000 to $10,000. Build it, grow it on the format library, flip it.

**Done-for-you shorts **

Creators and small brands pay $80 to $250 per short. One intake form, 10 shorts a month on retainer clears $2,000 to $5,000. Pipeline cost per short: pennies.

**The research service **

The most valuable output is the swipe sheet itself. Sell the outlier tracker and format library as a monthly report. 40 subscribers at $29/month is $1,160/month for research you already do for yourself.

**Format and prompt packs **

Sell the Claude prompts and the Make scenario on Gumroad. "Viral Format Finder Kit" at $37. Sell indefinitely.

The Numbers

Month 1: setup, $0 Month 2: $500-1,100 Month 3: $1,800-3,200 Month 6: $4,500-7,000 Month 9: $7,000-10,000 Month 12: $11,000-16,000+

The research loop compounds. Every month the sheet gets sharper and the hit rate climbs.

The Cost

Claude Pro: $20/month CapCut Pro: about $15/month Make Core: $9/month Google Workspace: $6/month

Total: $50/month in. $11,000+ out.

The Bottom Line

Most people make the video first and hope it finds an audience.

This runs the other way. Find the format that is already winning, feed it a fresh subject, then build.

Four tools. One loop. The research does the heavy lifting.

comment "VIRAL" and I'll send the outlier tracker and the Make scenario.