$ cat prompt.txt — Ecommerce & Shopify
High-Converting Product Description — 3x Your Page Conversion Rate
Product description writer that adds problem-framing, social proof, and urgency — conversion-optimised.
Most Shopify product descriptions are spec sheets. They don't sell. This prompt produces a 5-section description structure (problem framing, transformation promise, feature-to-benefit translation, social proof, urgency close) calibrated to your brand voice and price point. Tested across DTC brands, this structure routinely lifts add-to-cart rates by 15–40%.
$ cat prompt.txt | pbcopy
The prompt — copy & paste
You are a senior DTC ecommerce copywriter who has written product descriptions for 8-figure Shopify brands. Your job is to write a HIGH-CONVERTING product description for the product below using a proven 5-section structure.
PRODUCT INPUTS:
- Product name: {{product_name}}
- Specs / features list: {{product_specs}}
- Target buyer (be specific — current frustrations, life context): {{target_buyer}}
- Price point: {{price_point}}
- Brand voice (3 adjectives — e.g. "calm, confident, slightly witty"): {{brand_voice}}
OUTPUT FORMAT — write exactly these 5 sections, in this order:
SECTION 1 — HOOK (1 short paragraph)
Open with a problem statement that makes the target buyer nod. Reference the specific frustration they have BEFORE buying this product. Don't mention the product yet — focus entirely on the pain.
SECTION 2 — TRANSFORMATION (1 paragraph)
Paint the after-state. What does the buyer's day / week / life look like with this problem solved? Be sensory and specific. Avoid generic adjectives like "amazing" or "premium".
SECTION 3 — FEATURE → BENEFIT TRANSLATION (bulleted list, 4–6 items)
For each major feature from the spec list, write it as a bullet in this format:
• [Benefit-led headline (5–8 words)] — [Feature mechanism] [Why it matters to the buyer]
Example: "• Sleep through delivery trucks — Triple-layer foam dampens external noise by 28dB, so you stop waking at 6am."
SECTION 4 — SOCIAL PROOF (2–3 sentences)
Reference category-level proof (industry stats, expert endorsements) and reader-relevant outcomes. If we don't have real testimonials yet, use category-validated phrasing — "the same foam used by [adjacent premium brand category]".
SECTION 5 — URGENCY / RISK-REVERSAL CLOSE (1 short paragraph)
End with a reason to buy NOW (limited release, fast-selling inventory, seasonal relevance — whichever is true) AND a risk reversal (return policy, guarantee, "try it 30 days"). One short close paragraph, not bullets.
REQUIREMENTS:
- Voice matches {{brand_voice}} consistently throughout
- Total length: 180–250 words (mobile readers won't scroll further)
- Use sentence fragments where appropriate. Punchy. Like this.
- ZERO marketing-speak ("best-in-class", "revolutionary", "next-generation")
- Use "you" / "your" (second person) throughout
- The first 12 words must include either the buyer's frustration OR the transformation — these are the only ones above the fold on mobile
After the description, output one bonus section:
• META DESCRIPTION (150–155 chars) — for the product page <meta description> tag, optimised for the most likely Google search query a buyer would use.
If any of my inputs are vague (especially target_buyer or brand_voice), ASK for specificity before writing. A weak description from weak inputs is worse than no description.$ variables_to_fill_in
$ man playbook
Step-by-step playbook
How to actually use this prompt for the best results.
- 01
Gather your product specs in advance
{{product_specs}} should include materials, dimensions, capabilities, what's in the box. The richer your input, the sharper the feature-to-benefit translation.
- 02
Define your brand voice in 3 words
Most brands don't have a documented voice. Pick 3 adjectives that capture the tone — 'calm, expert, slightly witty' or 'punchy, irreverent, no-bullshit'. Consistency across products matters more than the specific voice you pick.
- 03
Describe the target buyer's pre-purchase frustration
This is the highest-leverage input. 'Tired moms' is weak. 'Moms of toddlers who've tried 3 sleep solutions and still wake at 5am' is strong. Specificity here flows through every section.
- 04
Run the prompt and edit the hook
Section 1 is the only section a mobile buyer will read before scrolling. Even when the AI gets it 90% right, spending 2 minutes hand-editing the hook is the highest-ROI activity in this entire workflow.
- 05
A/B test against your current description
If your store does >50 sessions/day to the product page, A/B test the new copy vs the current one (Shopify apps like Intelligems make this 1-click). Expect 10–40% lift on well-targeted products.
$ man tips
Pro tips for better output
- ›Generate 3 versions and use Section 1 from the strongest one — hooks are the variance-driver, the rest of the structure is consistent.
- ›Re-run the prompt after launch with real customer-review themes added to {{target_buyer}} — second-pass descriptions outperform first-pass by another 10–20%.
- ›For accessory / cross-sell products, mention the hero product by name in Section 2 — it boosts add-on conversion 30%+.
- ›Replace AI-default words ('elevate', 'curated', 'crafted') in your final pass — they're a tell that lowers trust.
$ echo $YIELD
What you'll get
A 180–250 word product description structured in 5 conversion-optimised sections, plus a 150-character meta description for the product page <meta> tag.
$ man faq
FAQ
Does Google penalise AI-written product descriptions?
No. Google's policy is helpful-content-first: if the description genuinely helps a buyer make a purchase decision, it ranks. AI-assisted is fine; AI-generated boilerplate that says nothing useful is what gets demoted. The 5-section structure here forces specificity, which is what matters.
Can I use this for Amazon listings?
Adapt it — Amazon listings need bullet-led structure with the most important benefits in bullets 1–3, no HTML formatting, and keyword density in the back-end search terms. Use Section 3's benefit bullets and skip the SEO meta-description bonus.
$ ls /prompts/ecommerce
Related prompts
$ cd /prompts — explore all 48 free AI prompts on ClaudeSkill