YouTube keyword research without guessing

On 2026-08-23 we ran a keyword measurement across 60 YouTube niches and returned a single number for each: an opportunity score from 0 to 100, marked estimated=false, meaning it comes from observed data rather than a model's guess. The spread was narrow: every niche scored between 49 and 63. This article explains what that method measures, what it does not, and how to use it before you commit a channel to a topic.

The short answer

YouTube keyword research without guessing means replacing intuition with three measured values per niche: demand (search volume on a 0-100 scale), saturation (competition, where high is bad), and opportunity (overall, where high is good). On 2026-08-23, measuring 60 niches this way produced demand between 61 and 100, saturation between 69 and 100, and opportunity between 49 and 63. The top-scoring niches were motivational and inspirational at 63; the lowest was life pro tips at 49. The gap is small, so the numbers guide you rather than decide for you.

What the three scores actually mean

The engine returns three independent readings per search term:

Score	Scale	Direction
Demand	0-100	Higher means more search volume
Saturation	0-100	Higher means more crowded
Opportunity	0-100	Higher means better opening

Demand tells you people are searching. Saturation tells you how many established channels already serve them. Opportunity combines the two into one figure you can sort by. In our 2026-08-23 run of 60 niches, demand ranged from 61 (life pro tips) to 100 (natural disasters, heists and scams, gaming stories, sports stories, motivational, inspirational). Saturation ranged from 69 to 100. No niche scored below 49 on opportunity, which is itself a finding: at this sample size, the differences between niches are real but modest.

A score marked estimated=false is not a forecast. It is a reading taken at a point in time. Treat it like a thermometer, not a promise.

Why most niches cluster in the middle

The narrow 49-to-63 opportunity band has a simple cause: popular topics are popular because they work, which means they are also crowded. True crime scored demand 92 but saturation 94, giving opportunity 58. History documentaries scored demand 97 against saturation 100, giving opportunity 58. Aviation disasters paired demand 94 with the lowest saturation in the set at 85, producing one of the better openings at 62.

The pattern repeats. Cold war: demand 98, saturation 92, opportunity 62. Espionage: demand 83, saturation 91, opportunity 53. When demand is near its ceiling, saturation usually follows within a few points. What separates a good pick from a bad one is rarely raw demand; it is whether demand outruns saturation by enough to leave room.

How to read the table for your own decision

Sort candidates by opportunity first, then check the two components behind each number. A niche scoring well because of high demand alone will fight you for attention. Here are five from the 2026-08-23 run:

Niche	Demand	Saturation	Opportunity
Motivational	100	92	63
Inspirational	100	92	63
Aviation disasters	94	85	62
Sports stories	100	95	62
Islamic stories	92	83	62

Then look at what the top videos in that niche actually are. For aviation disasters, the three highest-viewed results included a Smithsonian Channel video at 10,982,012 views from a channel with 663,000 subscribers, and a Mayday: Air Disaster video at 5,111,699 views from 1,100,000 subscribers. High view counts relative to subscriber counts signal that the topic rewards content quality over channel size, which matters if you are starting from zero.

For your own situation, do this in order. List the niches you could credibly produce weekly. Measure each with the same three scores. Drop any where saturation exceeds demand. Among the rest, pick the highest opportunity score whose top videos you can match or beat on specificity, not production budget.

Where guessing still beats measurement

Scores describe niches, not videos. Within aviation disasters, individual uploads range from 1,482,608 views down to 430 views on the same topic area. The niche-level number cannot tell you which specific subject inside it will perform. That judgment stays yours.

Measurement also cannot see your own fit. A 63 opportunity score in motivational is worth nothing if you have no interest in writing motivational scripts week after week. Consistency is a production constraint the engine does not measure, and we have not measured how much it affects outcomes; we only know the scores, not the channels built on them.

Finally, these readings come from one snapshot on 2026-08-23. Niches shift. A re-run next quarter may reorder them.

What this does not tell you

This article describes a measurement method and one dataset from it. It does not tell you which video ideas will succeed inside a niche, because we have not measured idea-level performance. It does not tell you how quickly a new channel gains traction, because we did not track channels over time. It does not predict algorithm changes or shifts in viewer behavior after 2026-08-23. And it does not cover monetization: revenue per view is outside this dataset entirely. If your question is "which exact video should I make next," this method narrows the field; it does not answer the final question.

<!-- faq -->

Frequently asked questions

How often should I re-measure a niche? We do not have a measured answer for cadence, because we took a single snapshot on 2026-08-23. What we can say is that all 60 scores came from one run, and a single run gives you a comparison point, not a trend. If you want trends, repeat the same measurement on the same terms and compare runs rather than trusting memory of past numbers.

Is a higher demand score always better? No. Natural disasters scored demand 100 on 2026-08-23 but saturation 99, leaving opportunity at 60. Aviation disasters scored demand 94 against saturation 85 and reached 62. The pair that matters is the gap between demand and saturation, which is why the opportunity score exists as a separate reading instead of a demand ranking.

What does estimated=false change about how I use the number? It means the value comes from observed data rather than a modeled estimate. You can cite it as a measurement with a date attached. It does not make the number permanent: saturation in particular moves as competitors enter and leave. Use it as evidence for a decision made now, not as a permanent property of the niche.

Can I apply this method to keywords outside the 60 measured niches? The method generalizes; the specific scores do not. We measured those 60 terms on 2026-08-23 and can vouch for those readings only. For a term we have not measured, run the same three-score process yourself, or state plainly that it is unmeasured. Writing an unmeasured number as if it were a reading is the same failure as guessing.