Every feature in Tubetific is built around one principle: decisions based on real YouTube data, not trends and hunches. That principle applies just as strictly to this page — of the 20 named signals in Tubetific's taxonomy, 5 have a live, working endpoint you can run today. The other 15 are real, named, and on the roadmap, not vaporware, but they aren't live yet. This page tells you which is which, because a tool built on 'real data, not guesswork' has no business being vague about its own feature set.
Tubetific runs a background scanner against the real YouTube Data API — not a scrape, not a third-party estimate — on a recurring schedule (trending signals refresh roughly every 6 hours; niche trend tables update every 3 days; pre-peak detection runs every 12 hours). That scanned data feeds a taxonomy of 20 named signals organized into 6 categories: Sub-Niche Divergence, Timing & Velocity, Audience Crossover, Format & Platform Shifts, Monetization Intelligence, and Competitive Blind Spots. Each signal is a distinct detection method looking for a distinct kind of opportunity — a rising sub-niche, a channel that went quiet, a language nobody's serving, a niche whose RPM makes it worth the production time. As of today, 5 of the 20 have a live endpoint: Style/Fusion Divergence (the flagship of the Sub-Niche Divergence category), Language Gap Inside Niche, Pre-Peak Keyword Detection, Upload Vacuum Alert, and CPM Arbitrage. Alongside those 5, Tubetific also has 4 separately-built core modules that aren't part of the 20-signal taxonomy at all but are equally real and live: Outlier Detection (VSP ratio), Content Gap Analysis, RPM Intelligence, and Rank Tracker — plus SAGE, an AI research assistant grounded in your real channel data, and a genuine, pixel-measured Thumbnail A/B scoring tool (not an AI guess dressed up as a score).
9 total live, working capabilities today: Outlier Detection, Content Gap Analysis, RPM Intelligence, Rank Tracker, SAGE, Thumbnail A/B, plus 3 of the 20-signal taxonomy's flagship live signals (Sub-Niche Divergence, Language Gap, Upload Vacuum)
5 of 20 taxonomy signals are live: Style/Fusion Divergence, Language Gap Inside Niche, Pre-Peak Keyword Detection, Upload Vacuum Alert, CPM Arbitrage
15 of 20 taxonomy signals are real, named, and roadmapped — not yet live — including Audience Overlap Map, Format Shift Detector, Comment Demand Mining, and Rising Micro-Creator Alert
Every live signal is checked against the same rule: if real YouTube data isn't available, the analysis doesn't run — Claude (Anthropic) analyzes the data, it never fabricates it
120 niches tracked across 50 countries, refreshed on the schedule above, not on-demand scraping at request time
The scanner runs on a fixed schedule, not on-demand
Trending/outlier data refreshes roughly every 6 hours, niche trend tables every 3 days, pre-peak detection every 12 hours. This matters practically: a signal you check twice in the same hour won't have new data, but checking back after the next refresh window will.
Each signal has a distinct trigger condition
Outlier Detection triggers on VSP ratio (views far exceeding what a small channel's subscriber count predicts). Content Gap Analysis triggers on a mismatch between search demand and comment-mined viewer satisfaction. RPM Intelligence doesn't 'trigger' at all — it's a lookup against Tubetific's monetization database for a given niche. Each of the 20 taxonomy signals has its own similarly specific trigger — they aren't 20 variations on the same detection logic.
Small channels are weighted differently than large ones
Outlier Detection specifically only qualifies channels under roughly 5,000 subscribers — a large channel getting a lot of views is expected, not a signal. This same small-channel weighting logic underlies several of the 20-signal taxonomy's detection methods, since the entire point is finding what the algorithm is rewarding independent of an existing audience.
Live signals produce real output; roadmap signals don't
This is the distinction this page exists to make clearly. The 5 live taxonomy signals plus the 4 core modules return real, current data when you run them. The 15 roadmap signals are genuine product plans with a defined detection concept, but running them isn't possible yet because the endpoint doesn't exist. Both categories are described honestly on this page and on each signal's own page — never blurred.
SAGE reads across your real data, not a canned script
SAGE builds an overnight brain note summarizing what changed in your channel and niche while you were away, and answers questions grounded in that real data rather than generic YouTube advice. It's powered by Claude (Anthropic) and — unlike some of the signal roadmap items — fully live and working today, included on the Pro plan.
Example
How three live signals compound on a real content decision
RPM Intelligence check: personal finance is A-Tier ($8-18 RPM) — worth the production time before anything else is decided
Language Gap check (live signal): the specific finance sub-topic under consideration has strong English-language dominance in a market where the primary searchers aren't native English speakers — a real, currently-underserved angle
Content Gap Analysis check: existing top-ranking videos in that language/topic combination have below-average engagement relative to views, with comments repeating the same missing information
Result: a content decision backed by three independent live signals agreeing, not a single number treated as gospel — this is the intended workflow, not a hypothetical
Result: None of these three checks require a roadmap signal — Outlier Detection, RPM Intelligence, Language Gap, and Content Gap Analysis are all live today and can be run in this exact sequence right now.
| Tool | Shows WHY a video is an outlier | Content gap analysis from comments | RPM by niche before you make the video | AI research assistant included |
|---|---|---|---|---|
| Tubetific | Yes — VSP ratio, live | Yes — live | Yes — live, all plans | Yes — SAGE, live, all plans |
| Other trend/score browser extensions | Partial — general trend/score metrics, not a small-channel VSP signal | No dedicated comment-gap tool | Limited — general monetization scoring, not niche-specific RPM ranges | No |
| Tag and keyword optimization tools | No dedicated outlier detection | No | No | No |
| Channel growth/rank tracking tools | No — tracks growth/rank history, not outlier detection | No | No | No |
| Native YouTube Studio Analytics | No — shows your own video performance only, not market-wide outliers | No | No | No |
Outlier Detection
Live. Finds videos where small channels (under ~5,000 subs) get views far exceeding their normal reach — the VSP ratio. Reveals what the algorithm is actively pushing right now, not what worked historically.
Content Gap Analysis
Live. Mines comments on top-ranking videos for the specific missing information viewers keep asking for — high search demand with unsatisfied viewers is the actual gap, not just high search volume alone.
RPM Intelligence
Live. Real RPM ranges by niche — e.g. finance $8-18, entertainment closer to $1 — so you know the monetization ceiling before investing production time, not after uploading.
Rank Tracker
Live. Tracks where your own videos and target keywords rank over time, so you can see whether a specific piece of content is actually gaining ground, not just guess.
Sub-Niche Divergence (Style/Fusion Divergence)
Live — one of the 20-signal taxonomy's 5 built signals. Detects when a broad category is declining but a specific sub-niche inside it is actually rising, before that sub-niche has a name competitors are searching for yet.
Language Gap Inside Niche
Live — another of the 5 built taxonomy signals. Flags topics where the top-ranking content is dominated by one language while a different-language audience is actively searching and underserved.
Pre-Peak Keyword Detection
Live. Flags topics with active, rising search activity where the current top-ranking videos are stale (2+ years old) — an imminent content gap, not a saturated one.
Upload Vacuum Alert
Live. Finds channels that have gone quiet (30+ days without uploading) while their audience is still actively searching for that content — an open window, not permanent competition.
SAGE AI Research Assistant
Live and included on the Pro plan. Grounded in your real channel and niche data, not generic advice — builds an overnight brain note and answers direct questions, powered by Claude (Anthropic).
Thumbnail A/B (pixel analysis)
Live. Measures real pixel statistics — brightness, contrast, saturation, warmth, edge density — on your thumbnail against real reference thumbnails from top-performing videos in your niche. A measured comparison, not an AI guess dressed up as a score.
CPM Arbitrage
Live — the 5th of the 20-signal taxonomy's built signals, part of the Monetization Intelligence category. Surfaces language/region combinations with real advertiser demand (high CPM) and near-zero existing supply.
What's still roadmap (15 of 20 taxonomy signals)
Real, named, and planned — not yet live: Diaspora Content Vacuum, Regional Flavor Crossover, Seasonal Anticipation, Shorts-to-Long Migration, Audience Overlap Map, Mood-to-Niche Mapping, Collaborator Network, Format Shift Detector, Thumbnail Style Shift, Title Tone Shift, Brand Gap, Underpriced Audience, Comment Demand Mining, Title Exhaustion, and Rising Micro-Creator Alert. Each has a defined detection concept; none has a live endpoint yet.
Does Tubetific use real YouTube data?
Yes. Every live signal pulls from the real YouTube Data API, not a scrape or third-party estimate. Claude (Anthropic) analyzes the data but never fabricates it — if real YouTube data isn't available for a given check, the analysis doesn't run rather than returning an invented result.
How often is the data updated?
Trending and outlier data refreshes roughly every 6 hours. Niche trend tables update every 3 days. Pre-peak detection runs every 12 hours. RPM Intelligence is a database lookup rather than a live scan, updated as Tubetific's monetization data is refreshed.
What does it mean that only 5 of the 20 signals are 'live'?
Tubetific's signal taxonomy names 20 distinct detection methods across 6 categories. 5 of them — Style/Fusion Divergence, Language Gap Inside Niche, Pre-Peak Keyword Detection, Upload Vacuum Alert, and CPM Arbitrage — have a real, working endpoint you can run today. The other 15 are genuine product plans, not fabricated placeholders, but they don't have a live endpoint yet — this page and each signal's own page state that distinction directly rather than blurring it.
Is Tubetific different from other YouTube research tools?
Most popular YouTube browser extensions and research tools are primarily optimization-after-the-fact tools — keyword and tag scoring for a video you've already decided to make. Tubetific's core differentiator is deciding what to make before investing production time: RPM Intelligence and Content Gap Analysis specifically address that earlier decision point, which is a real, distinct gap from what those tools generally cover.
Is SAGE available on Free or Plus?
No — SAGE is included on the Pro plan ($49/mo) specifically. Free and Plus include the full set of live research modules (Outlier Detection, Content Gap Analysis, RPM Intelligence, Rank Tracker, and the 5 live signal-taxonomy signals) at their own daily limits; SAGE's always-on AI research layer is the Pro-specific addition on top of that.
Is the Thumbnail A/B tool an AI guess or a real measurement?
It's a real, measured comparison — actual pixel statistics (brightness, contrast, saturation, warmth, edge density) computed on your thumbnail and on real reference thumbnails from top-performing videos in your niche, then compared. It replaced an earlier version that used an AI-generated 0-100 score with no real computation behind it.
What's the most common misunderstanding about the '20 signals' claim?
Assuming all 20 are equally live and usable today. They're not — 5 have working endpoints, 15 are roadmap. Tubetific's own product positioning is built on real data over guesswork, so overstating the feature set to make the number sound bigger would directly contradict that positioning.
How do I know which specific signal to check first for a new video idea?
RPM Intelligence first (is the niche worth the time), then either Content Gap Analysis (is there an underserved angle) or Outlier Detection (what format is the algorithm currently rewarding) depending on whether you're starting from a topic or a format question.
Does Tubetific work for Shorts as well as long-form?
The live signals (Outlier Detection, Content Gap Analysis, RPM Intelligence, Rank Tracker, and the 5 live taxonomy signals) don't distinguish format in their core detection logic — they work against whatever content exists in a niche. Format-specific detection (Shorts-to-Long Migration is literally one of the named signals) is on the roadmap, not live yet.