How to Grow a YouTube Channel From Scratch: 6 Operator Tactics

Sam lists what people think grows a channel: editing, camera gear, lighting, hooks. then, on the captions: "the most important thing is topic." his first video over 100K views was a contrarian take on exactly that. his advice to anyone starting: "Literally go from zero followers and say a contrarian take in your market and get insane distribution." tactic 02 on the same page is his filter, the three T's: topic, title, thumbnail. "if I can't distill my idea into those three T's, then we're not going to make that video." and the team did not settle for the first idea: "We would come up with 10, 15, 20 ideas to make one video."

the tactic: spend your effort on what the video says. a sharp take shot on a phone beats a polished video nobody needed.

try this week: write down the most common belief in your niche. list ten ideas that push back on it with something you know. for each, write a topic, a title and a thumbnail idea in one line. only keep the ones that fit all three.

Arab's agency skips paid keyword tools. on the captions, you take each content pillar and "you just use the YouTube search bar." how: "put yourself in the prospect's shoes and just start typing things into the search bar naturally." then scroll and look at the views on what comes up. "if you see there's traffic and traction for those topics, then you're good to go." if every video on a topic flops, skip it. Trinder, who has a computer science background, takes the same idea further. you go as deep as you can on one search term, open the channels behind the outliers, then step back and take the next term. done right, on his captions, "you're able to effectively explore the entirety of a niche and not miss anything."

the tactic: let viewers tell you what to make. if nobody watches videos on a topic, yours will not change that.

try this week: type five problems your buyer has into the YouTube search bar, exactly as they would say them. note the suggestions and the top videos. mark any video with far more views than its channel has subscribers. those topics go on your list.

Goh's order is backwards on purpose. on the captions: "you have to start with the thumbnail and title before you even do the hook, before you even do the script, before you even have the idea." the question he starts from: "What is a thumbnail and title that your ICP would click on based on their pain points?" (ICP means your ideal customer). YouTube viewers choose: "they're intentionally clicking on your video because of the thumbnail and title." Jordan Welch, on another Trakyo Pod episode, made the same switch. on the captions: "I would create videos and then do the title and thumbnail after the fact. These days I do everything completely backwards." his summary: "the 8020 of YouTube is the packaging and the ideas themselves."

the tactic: package first, then make the video that fits the package. if the title and thumbnail do not work on paper, the video will not save them.

try this week: before you film anything, write five titles and sketch three thumbnails for your next video. show them to two people who fit your viewer and ask which one they would click. film the winner.

after the click, the viewer has questions: is this going to deliver, and is it for me? Trinder answers both in the first seconds. on the captions: "Promise, proof, plan, persona. Four Ps. That's the framework that I use for all hooks." the promise restates what the title said in new words, so they know they clicked the right video. the proof shows why you can be trusted. the plan tells them what is coming. the persona confirms the video is for someone like them. his bigger rule sits behind it: "YouTube first and foremost is 100% a quality game, not a quantity game." Goh's version of the same idea: answer the viewer's top objections in the opening, before they find a reason to leave.

the tactic: script the first 30 seconds as four beats. promise, proof, plan, persona, in that order.

try this week: write the opening of your next video as four sentences, one per P. read it out loud. cut anything that is not one of the four.

Goh's case for over-delivering is about the next video, not this one. on the captions, when you over-deliver "you are training them to constantly click on your videos." his illustration: "If you do a good job, 80 of them come back" out of 100 viewers, and the baseline stacks from there. a weak video trains people the other way. when a video does flop, Trinder looks at the shape of the drop, not one number. mostly, on his captions, "if you look at the retention graph not really the retention percentage." a strong hook followed by a steep slide right after it means the value came too late: "That's a sign of low time to value." a sharp drop at a hard call to action means cut it or soften it next time.

the tactic: deliver in full, then let the graph tell you where you broke the promise. views say what happened. the retention graph says why.

try this week: open the retention graph of your last three videos. mark the steepest drop in each and write down what was on screen at that second. fix that pattern in the next video.

Arab's first rule, on the captions: "YouTube is an ecosystem of videos." one video rarely sells: "before somebody converts, they're typically watching like four or five videos or more before they make a buying decision." his channels mix searchable top-of-funnel videos, deeper videos further down, and one free course video, all built to last: "Think of each video as like a property that you're just collecting rent from." Goh measures the result by sales. on his captions: "the videos that I do that get the least amount of views and the least amount of likes are always the ones that actually get me the most amount of sales." he also says "The leads from YouTube, 85% of them pay in full. On Instagram, 54%." (his own business, not a benchmark).

the tactic: plan the next click and track the sale. a video is a step in a funnel, not a lottery ticket.

try this week: add one question to your booking form or checkout: which video brought you here? list your last ten videos and write down where each one should send a viewer next. link the next video at the end, not a request to subscribe.