Choose a YouTube Niche With Comparable Evidence

Start from the question, not the subject

"Personal finance" is a subject. "How do I actually start investing when I have a small amount of money and no idea what I am doing" is a question. Subjects do not tell you what to film; questions do, and they also tell you who is meant to click.

A workable niche has a question that recurs with variations. The recurrence is what makes a channel rather than a video: each upload answers another version, and the audience that came for one has a reason to return for the next.

Test your candidate by writing three specific questions in the viewer’s own words. If the three sound like the same video described differently, you have one video and not a niche. If they are genuinely distinct but obviously related, that is the shape you want.

Find comparable channels, not aspirational ones

The instinct is to look at the biggest channel in a space. That tells you the ceiling, not whether you can enter. What you want is channels roughly at your size, or a little above it, that are being watched for this question right now.

If you can find several such channels and their recent videos are performing above their own medians, that is real evidence: the audience exists and is reachable at your scale. If every channel answering this question is enormous, the demand is real but the entry may require an angle those channels cannot serve, usually something narrower or more specific.

If you cannot find comparable channels at all, be careful. It occasionally means genuine whitespace. Much more often it means people are not searching for this, and the absence of supply is the evidence.

Qualify on what you can sustain

Production cost is a selection criterion, not an afterthought. A niche requiring location shoots, long edits or expensive research may be perfectly good and still be wrong for you, because the thing that compounds on YouTube is publishing repeatedly enough to learn.

Estimate honestly: how long does one upload take end to end, and can you produce that at a consistent rhythm alongside everything else in your life? Then halve your optimism, because first estimates almost always assume ideal weeks.

A niche you can sustain weekly will usually outperform a more attractive niche you can sustain monthly, because it gives you four times as many chances to learn what your audience responds to.

Know what will confuse your results

When the first uploads land, several things other than niche quality will move the numbers, and mistaking them for signal is the usual way people abandon a good niche or persist with a bad one.

Format is the big one: if you switch from long-form to Shorts at the same time as switching niche, you cannot attribute the result. Audience carryover is another: your existing subscribers may click the first video out of habit and not the second, which makes the first look better than the niche is. Seasonality matters for anything tied to a calendar.

And video age matters, so always compare at a consistent point.

Hold format constant while you test a niche. Change one thing at a time here for the same reason you would anywhere else.

Plan four uploads, and say what would stop you

One video cannot test a niche, because a single result cannot separate the niche from the title, the thumbnail, the day, or luck. Four comparable uploads can begin to.

Plan four that answer four distinct versions of your question, in the same format, at a consistent cadence. Write down before you start what result across those four would justify continuing, and what result would make you stop. The second is the one people skip, and it is why sunk cost quietly takes over.

Kolliq is built for exactly this shape of decision: record the expectation, compare each upload against your own baseline rather than someone else’s, and keep the conclusion where the next decision can use it.

Review it as a hypothesis, not an identity

The point of the four-upload plan is that it ends in a decision rather than a vibe. At the review date, compare the four against your own median at a consistent age and check them against what you wrote down at the start.

Three outcomes are worth naming. Clear improvement means continue and deepen. Clear underperformance across all four, with format held constant, is real evidence to stop, and stopping early is a win rather than a failure.

Mixed results usually mean the question was too broad, and the useful move is narrowing it rather than abandoning it.

Niches are revisable. The creators who compound are the ones who treat a niche as a bet with a review date, not as a personality they now have to defend.

Your next review checklist

• Write three distinct viewer questions in the candidate niche.
• Confirm they are related but not the same video restated.
• Find comparable channels near your size, not just the biggest.
• Check whether their recent uploads beat their own medians.
• Estimate production time honestly and test it against your real week.
• Hold format constant for the duration of the test.
• Plan four uploads answering four versions of the question.
• Write the continue rule and the stop rule before starting.
• Review against your own median at a consistent video age.