I am preparing a YouTube campaign.

AUDIENCE:
[Describe.]

CHANNEL PROMISE:
[What viewers should expect.]

PRIMARY OBJECTIVE:
[Discovery / education / trust / release support / traffic / signup / sale / affiliate action / test.]

RESOURCE / RELEASE / OFFER:
[Exact thing the campaign supports.]

CAMPAIGN MESSAGE:
[One stable idea.]

CURRENT CATALOG:
[List relevant Shorts, long-form videos, playlists, songs, articles or resources.]

AVAILABLE ASSETS:
[Footage, audio, images, research, demonstrations, screenshots.]

PRIMARY CTA:
[One next action.]

EVIDENCE SIGNAL:
[What I will use to judge the campaign.]

VOICE:
[How I communicate.]

CONSTRAINTS:
[Time, production capacity, policy, disclosure, rights, claims.]

PREVIOUS EVIDENCE:
[CTR, retention, watch time, comments, traffic, conversions.]

Do not generate content yet.

First:
1. Restate the brief.
2. Separate facts from assumptions.
3. Identify missing information.
4. Identify what should remain fixed.
5. Recommend the smallest useful YouTube campaign scope.

Wait until the brief is coherent before building the campaign.

Review my existing YouTube catalog for this campaign.

Classify each asset as:
- DISCOVERY
- DEPTH
- PROOF
- CONTINUATION
- OWNED DESTINATION
- NOT RELEVANT

Then identify:
1. Which existing video should become the campaign anchor.
2. Which existing Shorts can support it.
3. Which playlist or next-video path should follow.
4. What genuinely needs to be created.
5. What should NOT be created because the catalog already covers the job.

Prioritize catalog reuse and sequencing over unnecessary uploads.

Using the approved YouTube campaign brief and catalog:

Build one campaign anchor around:
[Long-form video / full song / tutorial / case study / story.]

Create only the supporting Shorts needed.

For each Short:
- job;
- hook;
- opening visual;
- what it proves or teases;
- which long-form asset it supports;
- what should remain fixed;
- evidence signal.

Then define:
- long-form opening;
- mid-video proof / payoff;
- next-video or playlist handoff;
- external CTA only if appropriate.

Do not make the Short a compressed duplicate of the long-form video.
Each format needs its own job.

Create 4 packaging variations for the SAME YouTube video.

Keep fixed:
- audience;
- video content;
- core promise;
- factual claims.

Change only:
[title framing / thumbnail concept].

For each variation:
1. Working title.
2. Thumbnail concept.
3. What promise the pair makes.
4. Why the intended audience might understand it.
5. What stays fixed.
6. What evidence should determine whether another test is justified.

Do not use fake urgency.
Do not exaggerate what the video delivers.
Do not change the video topic between versions.

Using the approved YouTube campaign brief:

Build a 14-day campaign.

Rules:
- Keep one campaign message stable.
- Mark every asset as SHORT, LONG-FORM, PLAYLIST / SERIES, or OWNED HANDOFF.
- Do not schedule uploads just to fill days.
- Identify the anchor video.
- Identify which existing catalog assets support the campaign.
- Mark TEST assets and name the one variable being changed.
- State title / thumbnail role.
- State the opening job.
- State the evidence signal.
- Include one midpoint review before planning the second half.
- Preserve enough consistency to learn.

Here is my YouTube video opening:
[Paste transcript / outline.]

Audience:
[Paste.]

Title:
[Paste.]

Thumbnail promise:
[Describe.]

Review only the first 30–60 seconds.

Identify:
- what promise the viewer expects;
- when the video begins delivering it;
- unnecessary setup;
- ambiguity;
- evidence / proof timing;
- where a viewer may reasonably leave.

Recommend one controlled opening change.

Do not rewrite the entire video unless the opening cannot support the actual promise.

Use this completed long-form video:
[Paste transcript / outline.]

Create up to 5 possible Shorts.

Each Short must have a different job:
- context;
- demonstration;
- surprising result;
- objection / response;
- proof / process.

For each:
1. Hook.
2. Opening visual.
3. 20–45 second structure.
4. What long-form section it connects to.
5. Why it stands alone.
6. Whether a direct CTA is necessary.
7. Evidence signal.

Do not cut random clips just because they are short.

Here are my relevant YouTube videos:
[Paste titles + brief descriptions.]

Build a viewer path based on need, not upload date.

Group into:
- BEGIN HERE
- NEXT DEPTH
- SPECIALIST / ADVANCED
- CASE STUDY / PROOF
- RELEASE / PROJECT

Recommend playlist names and the order inside each playlist.

Explain the viewer logic.

Do not create playlists that contain only one unrelated video.

Campaign objective:
[Paste.]

What stayed fixed:
[Paste.]

Variable tested:
[Paste.]

YouTube results:
[Impressions, CTR, early retention, watch behavior, next-video / playlist behavior, comments, traffic, conversions.]

Qualitative feedback:
[Paste.]

Separate:
1. OBSERVED FACTS
2. POSSIBLE EXPLANATIONS
3. UNKNOWNS

Then recommend:
KEEP
CHANGE
TEST
STOP
CONTINUE OBSERVING

Do not infer causality from one upload.
Do not treat views as the universal success metric.
Recommend one next controlled change.

Create a reusable YouTube campaign memory from this completed cycle.

Record:
- AUDIENCE LANGUAGE
- STRONG TITLE / THUMBNAIL PROMISES
- WEAK PACKAGING
- OPENING LEARNINGS
- SHORT VS LONG-FORM LEARNINGS
- RETENTION PATTERNS
- PLAYLIST / NEXT-VIDEO LEARNINGS
- CTA LEARNINGS
- TRAFFIC / CONVERSION LEARNINGS
- PRODUCTION COST
- REUSABLE ASSETS
- UNRESOLVED QUESTIONS
- NEXT TEST

Label each insight:
CONFIRMED
PROMISING
UNCLEAR

Do not convert one successful video into a universal YouTube rule.