A copy-and-run production system for turning customer evidence into creator content. This is a workflow with prompts and editable templates, not software that automatically connects to your inbox.
Build a traceable path from customer questions to approved creative and a weekly production schedule.
<aside> 🚀 🚀 FREE STRATEGY CALL Want to grow your product through viral reach? We help teams across iGaming, mobile apps, AI products, and SaaS build creator-led organic distribution generating tens of millions of views at CPMs starting from $0.03.
<aside> 🚀 Book a free strategy call
</aside>
On the call you'll receive:
Use a spreadsheet for the board and your preferred AI assistant for analysis. Export only material you are permitted to process; remove names, emails, order numbers and sensitive customer details. Keep the originals in your company workspace. Never paste customer credentials into a prompt.
For a first run, assemble 30–50 recent reviews, 10 anonymized support excerpts, product specifications and your returns policy. This is a practical starting batch, not a statistical sampling requirement. Keep each source ID and date so a reviewer can trace every claim.
Create these tabs: Evidence, Clusters, Angles, Briefs, Production, Results. The prompts below produce rows for those tabs. Human approval remains mandatory before filming or publishing.
GOAL: distinguish a purchase barrier from a motivation and from a product defect.
ACTION: import source rows with columns ID, date, channel, product, exact quote and context. Deduplicate reposts. Do not treat repeated copies of one review as independent evidence.
EXPECTED OUTPUT: one row per independently sourced statement, labelled objection, motivation, use case, defect or irrelevant.
COMMON MISTAKE: interpreting a shipping complaint as a reason to create a performance claim.
Paste this prompt with your anonymized packet:
Act as a customer-evidence analyst. Use only the attached source rows and product facts. For each statement, return source_id, exact_quote, category, purchase_stage, implied_question, evidence_limit and confidence (high/medium/low). Keep quotes exact. Separate what the customer said from your interpretation. Flag ambiguous rows instead of forcing a label. Do not infer demographic traits or medical needs. Do not count duplicate quotes twice. End with missing information and questions for the owner.
GOAL: group rows by the decision they affect, not by vague keywords.
ACTION: separate “too expensive” into upfront price, uncertain durability and unclear difference from alternatives where the evidence supports it. Keep the supporting IDs with each cluster.
EXPECTED OUTPUT: named clusters with a customer question, supporting IDs, number of independent sources, contradictions and product proof needed.
Group the evidence rows by the underlying purchase decision. Keep motivation and objection in separate columns. Return cluster_id, decision_question, motivation, objection, source_ids, unique_source_count, contradictory_evidence and proof_required. A cluster with one source is a hypothesis, not a trend. Do not invent counts. Suggest a narrow content angle for each cluster, clearly labelled as a proposed angle.