Twenty inputs, four signals, one decision
The brief is explicit that the inputs are not the product: do not present them separately, combine them into demand, momentum, opening and fit, and read the content opportunity from those four. This screen is the ledger of what each input became.
The four signals and their weights
Each input feeds one or two signals; each signal answers one question. The combined score behind every badge on the board weighs demand at 30%, momentum at 25%, opening at 25% and fit at 20%. Demand carries the most because a subject nobody asks for is not an opportunity, however open the lane.
Do people want this?
Is interest growing?
- 01Emerging topics
- 03Competitor content performance
- 08Conversation velocity
- 11Evergreen potential
- 12Publishing timing
- 16Cross-platform signalsnot in this run
- 20Opportunity decaypartial
Is there room to own it?
Can WatchMojo win with it?
All twenty, in the brief’s order
What the brief asked for, in its own words, and what this run did about it. The status column is the honest part: the four signals are only as trustworthy as this list says they are.
- 01Emerging topics
Subjects gaining momentum before they saturate. Rate of growth over raw popularity.
MeasuredSubject share of uploads and of comments in the last 14 days against the 60 days before, as a lift ratio.
MomentumSee it - 02Audience demand
What viewers request, ask, debate, complain is missing, or want next.
MeasuredComment classes: requests, 'where is X', order debates and questions, each counted per subject.
DemandSee it - 03Competitor content performance
What outperforms competitors' own benchmarks, not simply what has the most views.
MeasuredEvery video's views over its own channel's median for the same type, on videos at least three days old.
OpeningMomentumSee it - 04Content saturation
How many creators already cover a subject versus how much interest exists.
PartialSupply counted across the 15-channel competitor set. A YouTube-wide search pass per subject was written but not run: YouTube began bot-checking the collector after 1,700 requests.
OpeningSee it - 05WatchMojo historical performance
How similar subjects, franchises, formats and categories performed on the channel.
Measured329 main-channel uploads from 25 May to 27 August, each indexed against the channel median, aggregated by subject, vertical and format.
FitSee it - 06Format performance
Which approaches work around a subject: rankings, explainers, retrospectives, comparisons, reactions, compilations, profiles.
MeasuredTen format lanes matched on titles; median index per format, WatchMojo against the field.
FitOpeningSee it - 07Audience conversation themes
The major ideas inside comments, not just sentiment.
Measured13,584 screened comments tagged by subject and by ten classes from requests to nostalgia.
DemandSee it - 08Conversation velocity
How quickly mentions and engagement are increasing or declining.
MeasuredEvery comment carries its own timestamp. Arrivals per day are charted overall and per subject for the last 60 days.
MomentumSee it - 09Content gaps
Subjects with audience attention that WatchMojo has not covered recently or competitors underserve.
MeasuredDays since WatchMojo last covered each subject, against demand and competitor supply.
OpeningSee it - 10Follow-up and sequel demand
Existing content generating requests for updates, second parts, expanded lists.
MeasuredRequest-class comments ('part two', 'do a top 30', 'next time') counted under owned videos and per subject.
DemandSee it - 11Evergreen potential
Sustained interest versus a short trend window.
MeasuredShare of comments arriving more than 30 days after upload, on videos old enough to tell.
MomentumSee it - 12Publishing timing
Now, within days, later in the cycle, or evergreen.
MeasuredA suggested lane from momentum, velocity and late share, which the authored call may override with reasons.
MomentumOpeningSee it - 13Title and hook patterns
Language and structures associated with stronger performance for comparable content.
Measured26 structural title features, each with median index with and without, WatchMojo against the field.
FitSee it - 14Thumbnail patterns
Faces, composition, text and other visual characteristics correlated with performance.
Not in this runNeeds image analysis over the thumbnails. The brand teardown records the current house style (single high-contrast face, the mojo badge bottom-left, rarely any headline text); performance correlation was not measured.
FitSee it - 15Audience language
Terminology, phrases, jokes and arguments that recur organically.
MeasuredVerbatim comments per subject, with the recurring phrases surfaced by discovery over the corpus.
DemandSee it - 16Cross-platform signals
Subjects accelerating on Reddit, TikTok, search and elsewhere.
Not in this runYouTube only. Reddit's keyless feeds were measured on an earlier COIL engagement and yielded too little to be worth it; TikTok and search need paid access.
MomentumRecorded here - 17Audience overlap
What other subjects appear around people engaging with a topic.
MeasuredSubjects raised in the comments under videos about a different subject, counted as pairs.
DemandOpeningSee it - 18Competitor white space
What competitors are not making despite audience demand.
MeasuredDemand rank against supply rank per subject. High demand and low supply is the white space.
OpeningSee it - 19Performance anomalies
Content significantly outperforming a channel's usual numbers.
Measured329 videos in the corpus beat their own channel median by 2x or more, ranked by the multiple rather than by views.
OpeningFitSee it - 20Opportunity decay
How quickly an opportunity disappears as competitors publish and conversation slows.
PartialVisible within the run as falling upload and conversation lift (Brand New Day is the worked example). A true decay curve needs a second collection run.
MomentumOpeningSee it
What a second run adds
4 of the 20 inputs are short of measured. 2 of them wait on another collection or a fuller pass; 2 wait on a collector this run did not have.
- 04Content saturation
Supply counted across the 15-channel competitor set. A YouTube-wide search pass per subject was written but not run: YouTube began bot-checking the collector after 1,700 requests.
- 20Opportunity decay
Visible within the run as falling upload and conversation lift (Brand New Day is the worked example). A true decay curve needs a second collection run.
- 14Thumbnail patterns
Needs image analysis over the thumbnails. The brand teardown records the current house style (single high-contrast face, the mojo badge bottom-left, rarely any headline text); performance correlation was not measured.
- 16Cross-platform signals
YouTube only. Reddit's keyless feeds were measured on an earlier COIL engagement and yielded too little to be worth it; TikTok and search need paid access.
A second collection run turns momentum into week-over-week and decay into a curve. Today momentum compares the last 14 days against the 60 before, inside one collection; with two, every subject can be scored against its own earlier self, and the 329 videos that beat their channel median by 2x or more can be watched as they fade.
Every signal is a 0-100 rank among the 262 subjects scored in this run, not a percentage of anything: a demand of 90 means the subject sits near the top of this run for demand, and the number moves when the field changes even if the subject does not. The run is a single collection on 2026-08-27. Comments are the top comments under each of 255 videos, our own cap, never a video’s full thread, and 13,584 of the 15,432 collected survived screening to be counted. Competitor supply is counted across the 15 channels in the competitor set, not across YouTube as a whole. The source is YouTube only, and no thumbnails were analysed.