Find Video Ideas People Already Want

Ideas are lying around in public

"I chose my direction… but what videos do I actually make?"

Here's the module's opening good news: you never have to invent video ideas from nothing again. The internet is a permanent, public record of what your audience wants — what they search, what they ask under other people's videos, what they complain nobody explains properly. Idea research isn't creativity under pressure; it's *collection*. The creativity comes later, in how YOU answer the demand.

You already did the first pass without realizing it: your Lesson 2 niche research produced a search harvest and a Small Channel Proof log. Those weren't just evidence for a decision — every line in them is a video idea wearing evidence. Jade's harvest contained 'forgotten history mysteries' and 'shortest war in history'; that second one became her first video. Today you'll systematically mine four sources like that — and attach an evidence level to everything you find, so when it's time to choose what to make, the choosing is easy.

One promise-filter reminder before the hunt: every idea you collect still has to pass your one-sentence promise from Module 1. Demand for a topic outside your promise is someone else's opportunity. The filter keeps your channel coherent while the research keeps it wanted — you need both, and now you have both.

The four idea sources

Four places demand announces itself — work them in order, log everything in the worksheet:

Evidence levels: proven, promising, hunch

Not all ideas arrive equal, and pretending they do is how beginners spend twenty hours on a video nobody wanted. So every idea gets a one-letter tag as you log it:

P — PROVEN. Direct evidence: it's a search suggestion AND you've seen it work on a small channel, or multiple people are asking it in comments. These are your safest bets — demand has receipts. Priya's 'weekly grocery budget for a family of four': rich suggestions, two small-channel outliers, asked twice in comments she read. Proven.

p — PROMISING. Indirect evidence: adjacent topics work, the theme repeats, but you haven't seen this exact idea validated. Most good ideas live here. Jade's 'the war that lasted 38 minutes' — no small channel had done it, but 'strange history stories' as a category was proven all over her research.

h — HUNCH. No external evidence yet — just your instinct. Log it anyway (next section explains why), but tag it honestly. Hunches aren't bad; unlabeled hunches masquerading as safe bets are.

The tags do their real work later: your first videos should lean P and p, because as a new channel you want demand working FOR you while you're still learning production. Hunches earn their shots once the channel has rhythm — and Module 5 will teach you to test them cheaply. One more axis while you're tagging: trend vs. evergreen. Trend ideas (this month's news, a viral moment) can spike but expire; evergreen ideas (how-to, explainers, timeless stories) get searched year after year. A beginner channel wants mostly evergreen — your videos are your library, and evergreen books stay checked out.

Inspiration vs. copying — the line, drawn precisely

Researching other creators means looking at other creators' work, so let's draw the line once, precisely, and keep it forever — it protects your integrity AND your channel.

Free to learn from: topics ('small channels do well with restoration timelines'), questions the audience asks, formats and structures (a before/after reveal, a countdown), and patterns ('titles with specific numbers pull in this niche'). These are the shared language of the platform — studying them is literacy, not theft.

Never yours to take: someone's script or wording, their thumbnail design, their title copied verbatim, their specific creative expression — recreating their video beat-for-beat with the nouns swapped. That's not inspiration; it's laundering someone's work. It can also bite legally and algorithmically — but the deeper reason is the one Module 1 taught: your channel's only durable advantage is being yours. A copied channel competes with the original at a permanent disadvantage: the original got there first and means it.

The practical test when you're unsure: *could the creator you studied watch your video and feel studied — or robbed?* 'They made a restoration timeline video, so I'll make one about MY bike, my way' — studied; completely fair. 'I paused their video every 30 seconds and rebuilt it' — robbed. When Jade found a small channel doing well with 'strange history' listicles, she took the *category* and made 'the war that lasted 38 minutes' — a story that channel never touched, told in her voice, from her research. That's the whole craft: demand is shared; the answer is yours.

Capture first, judge later

Last habit of the lesson, and it's the one that feeds every other lesson in this module: when an idea arrives, capture it — don't audition it.

Here's why the order matters. Judging at capture-time kills ideas for the wrong reasons: you're tired, so everything sounds mediocre; you just watched a polished video, so your idea feels amateur by comparison; you can't instantly see the thumbnail, so you assume there isn't one. Meanwhile the idea itself might be P-grade demand. Capture and judgment are different mental jobs — done together, judgment always wins and the page stays empty. Done separately — capture everything now, score it in Lesson 202's system — both jobs get done well.

So the rule: every idea from today's research session, and every idea that ambushes you in the shower next Tuesday, goes onto the list in seconds — messy, unjudged, tagged with whatever evidence letter it honestly deserves. Marcus's research session produced 31 ideas, from P-grade ('carb cleaning for absolute beginners' — all four sources lit up) to pure hunch ('restoring my grandfather's toolbox' — zero external evidence, but his gut says the story matters). All 31 go on the page. Which ones get MADE — and in what order — is a different question entirely, and it's exactly what the next lesson's Idea Bank and scoring system answers.

By the end of today's exercise you'll have 20+ ideas with evidence tags. That page is the raw material for your entire first quarter of publishing — collected from the wild, honestly labeled, and waiting for a system.