7 YouTube Video Ideas That Actually Get Views (and Repeat)

Here’s the thing most “video ideas” lists miss: a topic doesn’t get views, a format does. The same subject dies as a rambling explainer and detonates as a tight immersive story. The formats actually worth your time share three traits. They’re repeatable, so you can run the same shape every week without reinventing anything. They hook on the title alone, before anyone has watched a second. And they can be made faceless, which means AI can carry most of the production while you stay anonymous and ship fast.

Below are the seven I keep coming back to, in no particular order. This isn’t a ranking and there’s no “best” one at the top; the right format depends entirely on your topic, not its place on a list. For each one you’ll get a real example, the reason it holds attention, how it goes faceless, and the mistake that quietly kills it.

How to actually come up with ideas: steal the format, swap the subject

Most people brainstorm topics, publish four videos, run dry, and quit. The fix is to stop hunting topics and start collecting formats. A format is a shape that works regardless of subject, so once you have one, you have a hundred videos.

Take “Your life as X.” That single shape becomes “Your life as a Roman legionary,” then “Your life as a deep-sea anglerfish,” then “Your life as a medieval plague doctor.” Same structure, same production process, same thumbnail template, new topic each week. You’re not creating from scratch every time. You’re pouring this week’s subject into a mold you already trust. Lock five or six of these and the blank page stops existing. That single mental shift is the difference between a channel that lasts and one that burns out by upload eight.

It’s also your insurance against how fast YouTube moves. What’s pulling views this month often isn’t what pulls them next month. A niche heats up and cools off, a trend burns out, the algorithm leans a different way, and a subject that printed views in spring goes quiet by fall. Topics and niches are weather; formats are closer to climate. A good format keeps working while the specific subjects that win inside it rotate constantly, so you bet on the shape and stay loose on the fill. None of the examples below are picks I’d promise will be hot when you read this, and that’s the point: don’t marry a niche, keep one eye on your own analytics, and lean into whichever format and subject are landing for you right now, because that answer keeps changing too.

The 7 formats

”Your life as X”

This is second-person immersion, and it’s the most underrated format on YouTube. You don’t describe a world to the viewer. You drop them inside it and narrate their day.

Picture the opening of “Your life as a Roman legionary.” It’s 58 BC. You’re nineteen, you’ve just signed twenty-five years of your life to the legions, and tomorrow you march twenty miles under sixty pounds of kit. From there the video lives a day in that skin: the food, the boredom, the terror of the front line, the friend who doesn’t make it home. The viewer isn’t watching a soldier. They are the soldier.

It works because second person plus present tense collapses the distance between viewer and subject. You physically cannot spectate when the script keeps saying “you.” That’s why retention on these runs so high. It’s also endlessly repeatable across any era, job, animal, or planet, and it’s pure narration over atmospheric images, which is exactly what AI visuals and a steady voice were built to make. The one mistake that breaks the spell: slipping into third person. The moment it becomes “the legionary would have felt,” the magic dies. Stay in “you.”

The rise and fall of X

A real narrative arc, and the format where storytelling does the heavy lifting. The climb, the peak everyone remembers, the fatal decision, the collapse.

Run “The rise and fall of Blockbuster.” You open not at the top but at the humble start, because the audience needs to watch the empire get built before they’ll care that it burns. Then the dominance: 9,000 stores, a Friday-night ritual for a generation. Then the turn that makes the comment section detonate, the meeting where Blockbuster passed on buying Netflix for fifty million dollars. Then the slow, ugly unraveling.

It works on dramatic irony. The audience already knows how it ends, so every good decision early in the story drips with doom they can see coming and the characters can’t. Every company, empire, athlete, app, and trend has this arc, so you’ll never run out. This is also where a tool built for narrative pays off. TubeGen’s Story Mode is designed to plan the three-act structure first instead of dumping a flat summary, which is the whole game in this format. The mistake to avoid: starting at the peak. Skip the rise and the fall means nothing.

The explainer (“How X actually works”)

Take one genuinely confusing thing and make it click. “How does the stock market actually work?” is a permanent search query, which is the entire point of this format. People will ask that exact question every year forever, so the video compounds traffic for years instead of spiking on day one and dying on day three.

The craft here is restraint. You answer one question in layers, each a little deeper than the last, and you end at the exact moment the viewer finally gets it. The temptation is to cram ten related concepts in. Don’t. One idea, fully landed, beats ten half-explained every time. Production is the simplest on this list: a clean script, narration, and a few clear visuals or diagrams. It’s the most evergreen format you can build a channel on.

The “what if”

Speculative scenarios that hijack curiosity by promising a payoff the viewer can’t get anywhere else. “What if you never slept again?” “What if the dinosaurs never died out?”

The structure is a premise followed relentlessly through its consequences, hour by hour or step by step, grounded in real science wherever it exists and reasoned speculation where it doesn’t. The “never sleep” video walks the body’s collapse in order: the first groggy day, the hallucinations, the organ failure, the point of no return. That grounding is what separates a great “what if” from a lazy one. Anchor it in real mechanism and it feels plausible and unsettling. Make it pure fantasy and it feels like nothing. The premises are infinite and the production is narration over generated visuals of the scenario, so it scales as fast as you can think of questions.

The scale comparison (“X compared”)

These exist to put the viewer’s sense of scale on the rack and break it. “Every object in the universe, smallest to largest.” “How rich is the richest man in history, actually?”

It’s a relentless zoom. Each step recalibrates the viewer’s mental model, and the build climbs toward something that genuinely stops them: the moment a human realizes how small Earth is, or how much a billion actually dwarfs a million. It works because the brain physically cannot resist updating itself, and the payoff is pure awe, which is one of the most shared emotions on the platform. It’s also almost entirely visual, which makes it ideal for generated and animated shots with a calm narrator on top. Run the same shape on any axis: size, wealth, speed, time, distance.

The entire history of X

One subject, told start to finish, at a sprint. “The entire history of Rome in 20 minutes.” The appeal is the sweep, momentum carrying the viewer era to era without ever bogging down.

It earns views two ways at once. The long runtime means high total watch time, which the algorithm loves, and “history of X” is a search people make for years, so it stacks evergreen traffic on top. Every topic has a history, so the format never dries up. Production is long-form narration with an image per beat, faceless by default. The trap is dwelling. Spend six minutes on one decade and you’ve killed the momentum that made the format work. Keep moving.

The superlative countdown (“The deadliest X in history”)

Yes, it’s a list. The superlative is what rescues it from being wallpaper. “Top 10 disasters” is something the viewer has scrolled past a thousand times. “The 7 deadliest single days in human history” is a promise they have to see paid off.

The countdown structure holds people to number one by design, and the escalating stakes pull them up the ladder. It’s also the easiest format to template at volume, narration plus a visual per entry, which is why nearly every faceless channel keeps one in rotation. The whole format lives or dies on the strength of the superlative in the title. Pick a ranking worth arguing about and the video half-writes itself.

The part nobody tells you: a format is a series, and a series lives on its thumbnail

Every format above is really a series, not a video, and a series sinks or swims on whether people recognize it in a crowded feed. The feed is a wall of thumbnails competing for half a second of attention, and recognition is what earns the click. When a viewer has enjoyed two of your videos, the third only gets watched if they can spot that it’s yours at a glance. That glance is almost entirely thumbnail.

The faceless channels that blow up understand this and never design a thumbnail from scratch. They lock one template, one bold color, one text placement, one visual motif, and pour every episode into the same frame. By the fortieth upload, viewers identify the channel before they’ve read a word. That familiarity reads as trust, and trust is what gets the click in a feed full of strangers.

This is honestly the most useful thing I’ve found running channels through TubeGen. You set the thumbnail style once. Point it at a reference channel’s look, lock your colors and layout, and it templates that frame across the entire series, so episode 40 matches episode 1 without you opening a design tool forty times. Pair it with the rest of the pipeline, Story Mode for the narrative formats, a cloned voice for narration, generated images per scene, and the grind that ends most channels around upload ten simply doesn’t happen. The format is the idea. The templating is what makes running it forty times survivable.

The best format to start with

If you’re early, take “Your life as X” or a straight explainer. Both are narration over images, both are forgiving of a rough first attempt, and both let AI shoulder the production while you focus on the only thing that matters at the start: learning what your specific audience responds to. Your first video’s job isn’t to go viral. It’s to exist and generate real analytics, so your second video is built on data instead of guesses.

Pick one format, lock a thumbnail template, and ship episode one this week. Build your first one with TubeGen →