MOMENTUM / SIGNAL METHODOLOGY

What we know, what we infer, and what we never pretend.

MOMENTUM is useful only if its evidence boundaries are visible. This page explains the product's discovery, ranking, confidence, and AI grounding model.

Scan signals

What we measure

MOMENTUM retrieves public YouTube candidates, enriches them with source details, normalizes duration and statistics, classifies format and taxonomy, then ranks the matched sample.

retrieved candidatesenriched video detailsmatched filtersshown results

How momentum works

Momentum is not popularity. It weighs views per hour, freshness, engagement, reach, and evidence confidence. Raw views alone should not dominate the ranking.

velocityfreshnessengagementconfidence

What coverage means

A scan only represents the candidate pool MOMENTUM could retrieve and inspect. The UI should say top signals from candidates analyzed, not the best videos on YouTube.

coverage confidenceranking scoperetrieved atsource requests

What AI does

AI explains source evidence and helps create original concepts. It does not calculate views, rankings, velocity, or guaranteed performance.

grounded interpretationcreator opportunitiesrisk checksno invented data

Coverage language

Use honest product claims.

Say

Top MOMENTUM signals from the candidates analyzed.

Based on the latest India discovery sample.

Momentum ranking within retrieved public signals.

Do not say

The best 10 videos on YouTube.

Everything trending in India.

YouTube's official ranking.