youtube-content-gap-analysis

Map observed YouTube topic and format coverage using Crawlora searches, channel samples, videos, and available captions. Use for an evidence-based content brief or sampled topic-gap map with search selection, language, transcript basis, identity, and demand limits preserved.

Access and scope

Set CRAWLORA_API_KEY to your Crawlora key. Prefer the listed Crawlora MCP tools when connected; otherwise use the bundled scripts/crawlora.sh, which sends x-api-key to https://api.crawlora.net/api/v1. Read reference/endpoints.md for exact methods and parameters. Check application code as well as HTTP status; payloads are inside data. Stop on authentication errors, back off on 429, and retry a transient upstream failure once. Bound calls/pages by the requested scope and credit budget. Retain source IDs, URLs, source dates, and observation/crawl times separately.

Build a content-planning map for specified topics, audience needs, languages, markets, formats and channels. Define query synonyms, date window and balanced video/channel budgets. Search popularity is a discovery surface, not a complete inventory of every video or a measurement of unmet audience demand.

Collect a defensible video cohort

1. Use documented YouTube query/type/upload-date/duration/locale filters and returned continuation tokens. Shorts, long-form videos, livestreams and playlists are different units; keep the requested formats separate. Save every query, ordering, filter, sampled page and stopping reason.
2. Resolve actual video/channel IDs and canonical URLs. Channel names, handles, playlists and videos are different namespaces. Deduplicate the same video across searches; cross-channel reuploads need corroboration and remain distinct source observations rather than automatically independent content.
3. Inspect selected video metadata and channel archives. Record dates, duration, supplied views/engagement, creator, and coverage basis. Missing/rounded public fields stay unknown; no historical view curve is promised by this surface.
4. Discover transcript languages before requesting captions. Retain generated, original/translated language and timestamp context. Use JSON format for the normal envelope; text, srt, vtt return raw text. A transcript 404 means captions currently unavailable; a transient upstream error is different. Without captions/content evidence, label topic coding as title/description-only.

sh

scripts/crawlora.sh /youtube/search q="home energy efficiency" type=video upload_date=this_year sort_by=relevance
# Use actual returned video IDs before metadata and caption discovery:
# scripts/crawlora.sh "/youtube/transcript/$VIDEO_ID/languages"
# scripts/crawlora.sh "/youtube/transcript/$VIDEO_ID" format=json

Identify sample gaps and testable content hypotheses

Create a shared topic/subtopic, format, depth and intended-use rubric. Allow multi-label and uncertain states; cite short timestamped evidence only when the actual caption supports it. Automatic captions/translation can misrecognize speech and do not independently verify a claim. Do not reproduce full transcripts.

Return a topic-by-format/channel matrix with unique-video denominators, dates, collection scope and unavailable-content counts, then a prioritized idea brief linked to actual sample gaps. Low observed coverage does not prove a topic is absent, underserved or commercially attractive. High views or comments do not prove audience demographics, conversion, revenue, retention or replicable future performance. A causal/title/thumbnail recommendation needs further testing; label it a hypothesis rather than a guaranteed growth tactic. Do not upload, comment, message creators, download restricted media, or create ongoing monitors.