Seven levels of YouTube strategy: from reacting to every view count to building a channel that keeps winning

The video describes seven levels of YouTube strategy. The host says level two is where most creators get stuck and that level seven is essentially MrBeast. The framework is compared to Maslow's hierarchy of needs: at the bottom, a creator is just trying to survive the next upload, and at the top, they are building a channel that can win again and again.

The host stresses that the level a creator is on has nothing to do with their subscriber count. What it decides is the next video they try to post. The video says it will use real creators and patterns from VidIQ's data to show what each level looks like, where creators might be stuck, and how they can move up.

Level one is illustrated with a creator who checks views again five minutes after the last check, changes a thumbnail within an hour, and vows never to make a similar video after a flop. Seeing another creator's video in the same topic reach 100,000 views can then prompt a copycat decision. The host calls this tactical hell: decisions driven by the most recent result, which feel productive but do not move the channel forward.

VidIQ data is cited to show the behavior is more common on small channels. Channels under 1,000 subscribers change a video's title or thumbnail in the first three days about twice as often as channels with 10,000 to 100,000 subscribers. The host adds that this does not mean small channels are worse.

An extreme example is a channel with about 1,500 subscribers that made 10 title changes and nine thumbnail changes in its first 40 hours. The video passed 20,000 views, but the host asks which of the 19 changes the creator could learn from and repeat. The data, the host says, is not helpful to anyone, least of all the creator who made the changes.

To leave level one, the host says a creator must be able to explain what is wrong before changing anything and how the change might help. Simply being nervous about low views is not a good enough reason.

The host says level two is where most creators get stuck, including many with millions of subscribers. At this level another creator's result decides the next video. The upgrade over level one is that a creator starts from a format that has already worked for someone else instead of inventing ideas from scratch. The host notes that the video itself uses that approach with the 'seven levels of' title format.

The format has worked across niches. A 'seven levels of worldbuilding' video got nearly a million views on a channel with about 22,000 subscribers. A 'seven levels of speedsters' video got 445,000 views on a channel with about 1,000 subscribers. Versions exist on deadly snakes, mathematicians and logical thinking.

The problem is expecting the format to do all the work. Four months later, the worldbuilding channel uploaded 'seven levels of worldbuilders', which got about 80,000 views. The host says that is still good, but far below the first video, and suggests part of the first video's appeal may have come from viewers simply seeing something new.

The format has spread quickly. VidIQ's database shows about 45 long-form uploads with 'seven levels of' in the title in the third quarter of 2024, 370 a year later, and more than 1,200 last quarter, covering topics such as G-Wagons, rum, perfume dupes and horror movies. To leave level two, a creator must explain why the format worked and whether it still applies to their video.

At level three, the host says the creator's own prediction decides the next video. When a video flops, reasons come easily: the thumbnail was too busy, the topic was too narrow, or it was published at the wrong time. The host asks whether the creator actually thought those things through before publishing, or whether they only sound convincing now.

The host also points to a VidIQ MCP tool that can run the same kind of data check used in the video, through Claude or ChatGPT.

The video gives an example: a creator feels the last few thumbnails were too busy and hard to read on a phone screen, so they try a simpler design, expecting it to be easier to understand and to earn more clicks. That leaves one clear thing to check: did more people click the new design than the old one on recent videos? YouTube Studio reports this as click-through rate.

If click-through rate rose, the host says the creator still has to ask whether the thumbnail caused it or whether the topic appealed to a wider audience.

The host says the biggest creators also keep checking these assumptions. Chucky Appleby, who makes thumbnails for MrBeast, believed Jimmy looked better with his mouth open, while Jimmy preferred it closed. Rather than argue, they tested the question across 30 different videos.

The closed-mouth thumbnails earned higher watch time in every one of the 30 tests. Jimmy said he had known it. The host draws the lesson that if someone with that much experience can be wrong 30 times in a row, one upload and one test settle nothing. The suggested habit is to record a prediction and then the result over the next six uploads.

At level four, the creator studies what the biggest channels in their niche keep doing to win, rather than copying individual hits. The host uses Gabe Bott as an example. In the data, he had around 58,000 subscribers and his normal long-form videos got about 11,000 views.

In June he uploaded 'If You Have a Job, Start a YouTube Channel ASAP', which got over 460,000 views. Three months later, 'You Don't Need a Job to Make Money' passed half a million views. Those two videos accounted for 44% of the views across his last 49 long-form uploads.

The host argues the pattern is about the audience: people who are unhappy with or worried about their job and want other options. A creator can find this in their own niche by asking an AI tool with the VidIQ MCP for the five biggest channels on a topic and their outperforming videos, then asking who each channel is talking to and what it keeps promising. Leaving level four means being able to say in one sentence what each main competitor relies on.

Level five means choosing the next upload based on what nobody else in the topic is currently offering. The host tells the story of Miyamoto Musashi facing Sasaki Kojiro, a swordsman whose unusually long blade let him strike first. Musashi carved a wooden oar even longer than the sword, removing the one advantage Kojiro relied on rather than trying to become a better swordsman. The host's translation: find what everyone in the niche relies on and make something that makes it matter less.

The host warns this only works when videos in a topic look and sound alike. If they do not, doing something different may just confuse viewers.

To illustrate how crowded the grid has become, VidIQ pulled 100 recent breakout videos from 35 small channels teaching YouTube, all under 100,000 subscribers. Most of them relied on the same visual and title habits: faces, big text, red and green elements, and titles built around phrases like 'how to', 'why your' or 'settings you should change'. Only seven of the 100 opened with 'I did something'. Kevin Chee's earlier videos followed this same pattern, and across eight uploads a typical video got about 600 views.

Kevin Chee's first big jump came from personal content. After five years on YouTube, a video in which he said he was giving up got 24,000 views, about 40 times his previous average. He then leaned further into his own experiences: a 90-day consistency test brought nearly 138,000 views, over 200 times what his videos used to get. A video in which he used YouTube's own AI to solve the algorithm got 63,000 views.

The narrator says his thumbnails changed too, but the main strategy was going from telling people what to do, like everyone else, to showing how he did it. That made viewers want to see how it turned out for him, which the host says is how to leave level five.

The narrator asks viewers what they can show from their own experience that nobody else in their topic is showing. The answer could be an experiment, a project, or something that went badly. Instead of only talking about a topic, a creator can immerse themselves in it and make a video about that experience.

Level six is reached when a creator can say what every channel in their topic is doing and how they differ. Up to level five the focus was on making better videos; at level six it becomes what the whole body of uploads is building toward. The narrator says the most useful video is sometimes not the most-viewed one, and that sometimes you have to lose a battle to win the war. A creator may reach fewer people with one video in exchange for giving regular viewers a reason to come back.

Owen Co Games is a small team making an indie game called Tether Geist. Its channel has two kinds of videos. The first is video essays, such as one on why so many indie games use pixel art, which got over 100,000 views, and another that got almost 150,000 views. The narrator notes these subjects can be enjoyed even by people who have never heard of the game.

The second kind is devlogs. The channel has 41 devlogs showing how the game is being built. Jordan says these usually get fewer views than the essays, but they keep existing subscribers engaged and coming back, turning development into a series the audience wants to follow. Over time, the team built a Discord with over 1,200 members and fully funded the game on Kickstarter.

Before deciding a video is a disappointment, the narrator says to go back to why it was made. Was it meant to reach new people, help regular viewers with something specific, or build interest in a larger project? The narrator says a creator reaching level six can explain how the next few uploads connect and why someone who watches one would want to return for the next, planning something viewers can follow as well as something they can click.

At level seven, the goal is an audience that returns for whatever a creator offers, even if the format changes, along with a process that lets them repeat their best work. The narrator asks whether a channel would be fine if its best format stopped working tomorrow. MrBeast is placed at the top of the ladder, but the narrator says you do not need a channel anywhere near that size to start answering the question.

The narrator uses a car channel with about 119,000 subscribers at the time of recording to show what happens when a reliable format slows. In 2023, Joel's typical brutally honest review got nearly 200,000 views. Videos in that category are now down to 34,000 views, a drop of nearly 80%.

Despite that, his typical views this year are well above 2023 because he started other formats. In one, he buys cheap cars with problems and shows what needs fixing; 11 of those did far better than his usual videos. Another looks at why certain cars have become so cheap; the seven videos in that group were all breakouts. The narrator concludes that his biggest format lost 80% of its views while the channel grew anyway, which is how a creator keeps the channel running by finding new formats before the old ones run out.

The narrator lists the signs of level seven: you can explain how you will make your next strong video without relying on the last hit happening again, you know who you are serving, and you have a repeatable process you can improve. The aim is to find new ideas and formats before the old ones run out, so they are not propping up the entire channel.

To find your current level, start at level one and work up. The first level you are not ready to move on from is your level. The narrator notes this can be confusing: someone might plan a full series like a level six while still changing thumbnails every hour, as at level one. Shorts can look different, with creators seeing views stall at 1,000, 10,000 or 30,000, which the narrator calls view jails. The narrator asks viewers to say in the comments which level they are working on next.