Socialseal Video Concepting

SocialSeal Video Concepting

Overview

Video concepting is the ideation step that frames an opportunity into shootable routes and, crucially, into the scope or retrievalPrompt that drives the blueprint engine. A good concept names the search gap, picks a lens, and specifies the exemplar direction so socialseal-blueprint-builder can ground it in real videos.

See references/production-pipeline.md and references/content-lenses.md. Concepting precedes reference-video analysis and blueprint generation; it does not replace them. Concepts are hypotheses to ground in real surfacing videos and then test, not proven winners; see references/evidence-and-confidence.md.

For work that will proceed into production, read references/creative-production-gates.md. Concepts are a decision gate, not an excuse to select one route silently and build a large batch.

When to Use

• Translating an opportunity/gap into 3+ distinct concept families.
• Defining the blueprint scope (topic/competitor/tracking_group) or a retrievalPrompt per concept.
• Preparing concepts for grounding before any brief or externally edited rough cut.

Inputs

• target platform, market, and keyword/topic, with the opportunityKey
• opportunity analysis or competitor/content evidence
• brand context and constraints
• brand utility bank: real facts, tips, proof, exclusions, and facts needing confirmation
• available assets, creator type, or shoot context

Workflow

1. Name the search gap. One sentence: "People search for X, but current surfaced content lacks Y."
2. Select lens. Aspirational (mood/emotion) or utility/practical (teach/plan/compare).
3. Generate concept families. At least three distinct: tutorial/walkthrough, comparison, first-timer guide, POV/day-in-life, list/map/pin, mistake/avoidance, or mood montage.
4. For each concept, specify: hook, first frame, arc, hero shot, supporting shots, useful detail or mood cue, CTA, and target keyword/topic.
5. Attach an engine handle. For each concept, define the blueprint input: a scopeType (+ scope fields) or a retrievalPrompt that would retrieve matching exemplars. This is what makes the concept groundable.
6. Rank concepts. Score by evidence fit, usefulness/emotional pull, production ease, and measurement clarity.
7. Present the concept decision. Show the top 2-3 routes with trade-offs and ask the user to select, combine, or adjust them. Record the approved route and facts; do not produce every route unless requested.
8. Hand off. Send approved concepts to socialseal-reference-video-analysis (preview exemplars) and socialseal-blueprint-builder (generate).

Output

For each concept:

• name, viewer job, lens
• hook options and first 3 seconds
• arc / beats, hero shot, supporting shots
• caption direction and measurement keyword/topic
• blueprint input: scope or retrievalPrompt
• production difficulty
• brand utility/proof used and approval status

Do / Don't

Do:

• produce multiple directions, not one early "best" idea
• give every concept a concrete blueprint scope or retrievalPrompt
• connect every concept to a measurement signal
• keep concepts shootable with available assets

Don't:

• generate generic viral ideas detached from the keyword gap
• treat a concept as final content; it must be grounded by a blueprint first
• copy competitor hooks directly
• use scripted ad language

Troubleshooting

• Concepts feel generic: add the exact keyword and viewer hesitation, and a sharper retrievalPrompt.
• No exemplars surface in preview: widen the scope/keywords or revise the prompt before briefing.
• Concept does not map to a tracking group: add tracking coverage (socialseal-tracking-group-design) so it is measurable.

Verification Checklist

• Each concept has a viewer job and lens.
• Each concept has hook, first frame, arc, and hero shot.
• Each concept carries a blueprint scope or retrievalPrompt.
• Each concept maps to a measurement keyword/topic.
• Top concepts are ranked and ready for grounding.