Content Win Loss Reviewer

Analyze ecommerce or creator content post-launch to diagnose why it won or lost using evidence, scoring, and actionable lessons for improvement.

Review a piece of ecommerce or creator content after it runs and explain why it likely won or lost, using evidence, simple scoring, and actionable lessons for the next iteration.

Use this skill when a user wants a postmortem on a script, ad, creator video, landing asset, or social post. It is useful for separating surface-level reactions from operational lessons about hook, proof, offer, fit, execution, and distribution context.

Solves

Teams often say content “worked” or “flopped” without learning much:

• they over-credit views while ignoring commercial outcome;
• they blame the creator when the offer was weak;
• they blame the hook when retention was fine but CTA failed;
• they copy winners without understanding what really drove the result.

Goal: Turn a content result into a simple win/loss diagnosis with evidence, confidence level, and next-step recommendations.

Use when

• Reviewing a published creator post, ad, script, or content experiment
• Running postmortems after a launch, campaign, or test batch
• Comparing why one piece outperformed another
• Distilling lessons from wins without blindly copying them
• Distilling lessons from losses without vague blame

Do not use when

• There is no performance signal, observation, or content context to review
• The user needs statistical attribution modeling or media mix analysis
• The task is purely to rewrite copy without analysis

Inputs

• Content asset, transcript, script, or summary
• Observed outcome metrics or directional results
• Goal / KPI used to judge success
• Audience and channel context
• Product and offer details
• Distribution conditions (timing, spend, creator, traffic source)
• Comparison asset if available
• Known anomalies or confounders

Workflow

1. Define the success standard for this content.
2. Summarize the observed result and relevant context.
3. Break the outcome into likely drivers and likely blockers.
4. Score confidence for each explanation based on evidence quality.
5. Extract repeatable lessons and caution flags.
6. Recommend what to keep, change, retest, or stop.

Review dimensions

Use simple labels such as strong / mixed / weak or 1-5 scoring across:

• Hook / stopping power
• Message clarity
• Product relevance
• Proof / trust
• Offer strength
• CTA / next-step clarity
• Audience-content fit
• Distribution fit
• Learning confidence

Output

Return:

1. Outcome summary
2. Win/loss verdict
3. Likely drivers
4. Likely blockers
5. Confidence notes
6. Next-test recommendations
7. Reusable lessons

Quality bar

• Separate outcome facts from interpretation
• Distinguish creative problems from offer, audience, or distribution problems
• Avoid false certainty when evidence is thin
• Focus on lessons that change the next decision
• Keep the review operator-useful, not abstract

Resource

See references/output-template.md.