Build Your Core Asset

Initial Setup Checklist

  1. Provide the seller with a simple spreadsheet to collect raw facts (materials, dimensions, variants).
  2. Feed the raw facts into AI with strict instructions to output a specific format.
  3. Extract the technical details into a bulleted list.
  4. Add required care or caution notes.
  5. Manually read the AI output line-by-line and delete any claims not present in the raw data.

Action Prompts

PROMPT 1: Draft the Description
Here are the raw facts for a handmade soy candle: Seller is Nisha Candles. Scent is lavender. 100g jar. Simple gift packaging. Do not invent certifications, health benefits, burn time, or reviews. Output the structured description now.
PROMPT 2: Factual Verification
Review the description you just generated. List every single claim made in the text, and confirm whether it was explicitly provided in my original raw facts.

Finished Example

This is an illustrative finished example for educational purposes.
Title: Lavender Soy Wax Jar Candle - 100g A simple, handmade lavender candle designed for desks, bedside tables, or small gifting. Poured into a reusable glass jar. **Specifications:** - Material: Soy wax and lavender oil - Size: 100g - Packaging: Simple gift box included **Care Instructions:** Place on a heat-safe surface away from drafts. Never leave burning unattended.

Common Mistakes

  • Inventing facts (AI hallucination)

    Impact: The AI invents a "60-hour burn time." A buyer purchases it, the candle burns for 20 hours, the buyer demands a refund, and the seller fires you.

    Correction: Strictly instruct the AI to only use provided facts, and manually proofread every single line against the raw data spreadsheet.

  • Writing massive blocks of text

    Impact: Shoppers scanning on mobile phones will not read large paragraphs, defeating the purpose of the formatting.

    Correction: Use short sentences. Force all technical details (size, weight, material) into bullet points.

Failure Recovery Warning: Diagnostic guidelines for formatting work.

Practice Execution

  1. Create a basic fact-collection spreadsheet with columns for Material, Size, Weight, and Care.
  2. Fill in fake, messy raw data for 5 fictional products.
  3. Use the AI prompt to draft structured descriptions for all 5.
  4. Format them in a Google Doc with clear bold headings.
  5. Manually read each one and delete any unsupported marketing claims the AI added.
Supplementary Methodology & Evidence
Methodology & Evidence: AI models are prone to adding generic marketing fluff (e.g., "high quality," "durable," "eco-friendly") when asked to write product descriptions. The critical step in this workflow is the manual human review to strip out those unverified claims.

Ready to execute?

This playbook explains the work. My Mission is where you DO and TRACK the work.

Start Building