Brief Paid Social Creative That Stops the Scroll

    You are a performance creative strategist. On automated platforms, targeting is largely out of your hands — creative is the lever that still moves results.

What I'm advertising and to whom: {{offering_and_audience}}
The problem it solves, in the customer's words: {{problem}}
Proof I have — results, reviews, demonstrations, before-and-after: {{proof}}
What's worked and failed before: {{creative_history}}
Production capability — can I shoot video, do I have customers who'd appear, what's my budget: {{production_capacity}}
Platform and placements: {{platform}}

Produce:

**The first three seconds.** For each concept, what happens immediately. This is the whole job — a viewer decides before any message lands. Effective openers: the problem shown rather than described, an unexpected visual, a direct statement of the outcome, a recognizable moment of frustration, or a result shown first. Never a logo, never a slow build, never a brand statement.

**Concepts.** Five to eight distinct ones, each with a different underlying idea rather than a different edit. Draw across the formats that reliably work: the problem-solution demonstration, the customer speaking in their own words, a direct comparison against the current way, a founder or expert explaining something, a listicle or tips format that earns attention before selling, an unpolished native-looking clip, and a static image with strong text. For each: the concept, the opening, the message, and why it might work for {{offering_and_audience}}.

**Hook variations.** For the strongest two concepts, five different opening lines or visuals each. Hook testing on a proven concept is usually higher-return than testing entirely new concepts, and it's much cheaper.

**Match production to reality.** Against {{production_capacity}}: which concepts I can actually make. Note that lower-production, native-looking creative frequently outperforms polished brand work on social placements, so limited capacity is less of a constraint than it seems.

**Format requirements.** For {{platform}}: aspect ratios, safe zones, length, and text placement. Design for sound-off viewing with captions, since most of the audience won't hear it — a concept that depends on audio is a concept that fails silently.

**Use the proof.** From {{proof}}: where real numbers, real customers, or a demonstration go. Showing the thing working outperforms claiming it works.

**What to avoid**, from {{creative_history}} plus the usual: stock footage that signals an ad instantly, dense text, a logo opening, corporate voiceover, and anything requiring context the viewer doesn't have.

**Volume expectations.** Set the realistic bar: creative fatigue drives costs up sharply — often a large increase in cost per thousand impressions once an audience has seen something too often — and fresh creative can bring costs back down within days. Fatigue arrives faster in some categories than others, so plan for a steady flow rather than a quarterly batch. Advertisers producing many new creatives per month meaningfully outperform those producing a handful. Given {{production_capacity}}, say what monthly volume is achievable and what that implies.

**Naming and tracking.** A convention that records the concept, hook, and format, so the test results can be analyzed by variable rather than by filename.