These 55 prompts are for people who run an online store and write their own product copy: the Shopify founder with a 300-row catalog, the Amazon seller fixing a suppressed listing, the Etsy maker who would rather be making. Most lists of ChatGPT prompts for product descriptions give you a one-line request with a blank for the product name. These are complete briefs that cover the whole catalog job, from the product page to the abandoned cart email.
Each prompt gives the AI a role, asks for your real product data through {{VARIABLES}}, and sets rules that stop it inventing specs, certifications, ratings or "best seller" claims. Where a channel has rules, the prompt follows them: Amazon's title and bullet requirements, Etsy's 13 tags, eBay's 80-character titles, Google's structured data policies and email unsubscribe rules. Prompts for beauty, food and environmental copy flag claims that need evidence before they go live.
The pack covers descriptions by product type, marketplace listings, product and collection SEO, bulk workflows from a pasted spreadsheet, variants and size guides, review mining, email flows, ads and product photography, including AI image prompts for lifestyle shots. Four prompts include a worked example so you can see the output before you start. Every prompt works in ChatGPT, Claude and Gemini.
Product Descriptions by Product Type
Start with a brand voice card, then pick the prompt that matches what you sell: spec-heavy gear, apparel, beauty, food, handmade, digital downloads, or a reseller page that needs to stop reusing the manufacturer's text. Each prompt writes from the facts you paste and lists what's missing rather than inventing it. The beauty and food prompts keep cosmetic and food copy away from medical and nutrition claims that need evidence.
Best for: Keeping hundreds of product pages sounding like one brand
You are a senior e-commerce copy chief. Build a one-page brand voice card that I can paste at the top of every product-copy prompt so descriptions stay consistent across my catalog.
What we sell and why the brand exists: {{BRAND_DESCRIPTION}}
Who buys from us and what they care about: {{CUSTOMER}}
Two or three pieces of copy we like (ours, or a style we admire, pasted as text): {{SAMPLE_COPY}}
Words, phrases or habits we never want: {{WORDS_TO_AVOID}}
Produce:
1. Our voice in three adjectives, each with a "this, not that" pair (for example "confident, not boastful").
2. Sentence rules: typical sentence length, person (we/you), contractions or not, and how we use humor.
3. Vocabulary: 10 words we use and 10 we avoid. Include my list and generic filler.
4. A structure rule for a product page: what goes in the first line, the body and the closing line.
5. A 60-word sample description for a neutral product in our range, written in the voice.
6. A voice check: five yes/no questions to run on any draft.
Constraints: base the voice on the samples and the customer, not on a generic "luxury" or "fun" template. If the samples conflict with each other, point out the conflict and ask which to follow. Keep the whole card under 400 words.
Fill in: BRAND_DESCRIPTION CUSTOMER SAMPLE_COPY WORDS_TO_AVOID
Run this once, save the card, and paste it into any other prompt in this pack where it asks for brand voice.
Best for: Turning a supplier spec sheet into a scannable product page
You are a conversion copywriter for Shopify stores. Write the product page description for {{PRODUCT_NAME}}.
Source facts (spec sheet, supplier notes, my own notes): {{SPEC_SHEET}}
The problem or moment that makes someone buy this: {{CUSTOMER_PROBLEM}}
Brand voice: {{BRAND_VOICE}}
Target length: {{WORD_RANGE}} words
Structure:
1. Opening line (max 20 words): the main outcome for the buyer, specific to this product. No "Introducing", no questions, no exclamation marks.
2. A short paragraph (2–3 sentences) on who it's for and the moment it solves.
3. "Why you'll like it": 4–6 bullets, each pairing a feature from the source with what it means in use ("18/8 stainless steel, so it won't hold coffee smells").
4. "Details": a compact list of specs exactly as given (dimensions, materials, capacity, what's included).
5. One closing line that removes a hesitation using a fact from the source (warranty, returns, compatibility), only if I supplied it.
Use only the facts I've given: no invented specs, certifications, ratings, rankings or quotes. List anything missing under "Facts to confirm".
Then return the same description as clean HTML using only <p>, <ul>, <li> and <strong>, ready for the Shopify editor's HTML view, followed by the "Facts to confirm" list.
Fill in: PRODUCT_NAME SPEC_SHEET CUSTOMER_PROBLEM BRAND_VOICE WORD_RANGE
Paste the spec sheet as it is, including units. The HTML version pastes straight into the Shopify editor's code view.
Best for: Electronics, tools and equipment where buyers need the specs explained
You are a technical copywriter for electronics, tools and equipment stores. Write a description for {{PRODUCT_NAME}} that a buyer can understand without the manual and that an expert will trust.
Full specifications: {{SPECS}}
The buyer's level of expertise (beginner, enthusiast, trade): {{BUYER_LEVEL}}
Main jobs people buy it for: {{USE_CASES}}
What it works with and doesn't work with: {{COMPATIBILITY}}
Write:
1. A two-sentence summary of what it does and who it suits.
2. "What the specs mean": for the 5–7 specs that matter most for these use cases, one line each in the form spec → what it means in practice. Add a plain-language comparison only where the arithmetic follows from the specs I gave, and show the working.
3. A compatibility block: works with / doesn't work with / check before buying.
4. A "Not for you if…" line that sends the wrong buyer elsewhere, which cuts returns.
5. A full spec table in two columns using my exact units.
Rules: never round, convert or upgrade a spec silently; if you convert units, show both values. Don't use "best-in-class", "military-grade" or "industry-leading" unless I've supplied the test or source behind it. Use only the facts I've given: no invented specs, certifications, ratings, rankings or quotes. List anything missing under "Facts to confirm".
Fill in: PRODUCT_NAME SPECS BUYER_LEVEL USE_CASES COMPATIBILITY
Paste the full spec table. The 'Not for you if' line is there to cut returns from the wrong buyer.
Best for: Clothing pages that reduce 'not as expected' and sizing returns
You are a fashion e-commerce copywriter. Write the product description for this garment so shoppers can judge fit and feel without trying it on.
Garment and design details: {{GARMENT}}
Fabric composition and care label, exactly as on the label: {{FABRIC_AND_CARE}}
Fit notes (cut, true to size or not, what customers or our team said): {{FIT_NOTES}}
Model's height and the size worn, if photographed: {{MODEL_INFO}}
How our customers wear it: {{STYLING_CONTEXT}}
Output:
1. A 40–60 word description that leads with how it looks and feels on, using accurate sensory words (weight, drape, stretch, texture).
2. "Fit": 2–3 lines on cut, length and whether to size up or down, based only on my fit notes.
3. "Fabric & care": composition and care exactly as given, plus one line on how the fabric behaves, only if that is true of this fabric in general (for example, linen softening with washing).
4. A "Model is…" line.
5. Three styling suggestions using item types, not other brands' products.
Avoid "flattering on every body" and "fits everyone" unless it is a genuine one-size item. Don't call the fabric "sustainable", "organic" or "recycled" unless the composition or a certification I supplied supports it. Use only the facts I've given: no invented specs, certifications, ratings, rankings or quotes. List anything missing under "Facts to confirm".
Fill in: GARMENT FABRIC_AND_CARE FIT_NOTES MODEL_INFO STYLING_CONTEXT
Copy the fabric and care details straight from the garment label. Add fit notes from your own try-ons or return reasons.
Best for: Skincare, haircare and cosmetics pages that sell without drug claims
You are a beauty copywriter who knows the line between cosmetic claims and drug or medicinal claims. Write a product description for {{PRODUCT_NAME}}.
Full ingredient list (INCI): {{INGREDIENTS}}
What it does, in my words: {{WHAT_IT_DOES}}
Skin or hair types it's made for: {{SKIN_TYPES}}
Evidence I hold (consumer test results with sample sizes, lab tests, certifications): {{EVIDENCE}}
Where I sell (country or region): {{MARKET}}
Write:
1. A 100–150 word description focused on look, feel, texture, scent and how it fits a routine.
2. "Key ingredients": up to four, each with one sentence on its cosmetic role.
3. "How to use" in three steps.
4. A claims table: every claim in your copy, the evidence from my list that supports it, and a verdict of "OK", "needs evidence" or "remove".
Compliance rules: cosmetic claims describe appearance and feel ("looks smoother", "feels hydrated"). Don't claim to treat, cure, heal or prevent any condition (acne, eczema, hair loss, "repairs" wrinkles) or to change the body's structure, because in the US, UK and EU that can make it a drug or medicinal product. Quote consumer-test results only as I give them, with the sample size. No "clean", "non-toxic", "chemical-free" or "hypoallergenic" unless I supply the basis. Flag anything I should check with a regulatory adviser for {{MARKET}}.
Fill in: PRODUCT_NAME INGREDIENTS WHAT_IT_DOES SKIN_TYPES EVIDENCE MARKET
List every piece of evidence you actually hold, with sample sizes. The claims table shows which lines need proof before you publish.
Best for: Specialty food, drinks and pantry products with allergen accuracy
You are a food copywriter for an online specialty grocer. Write a mouth-watering but accurate description for {{PRODUCT_NAME}}.
Label text (ingredients, allergens, net weight, storage, nutrition panel): {{LABEL_TEXT}}
Our tasting notes: {{TASTE_NOTES}}
Producer or origin story (facts only): {{ORIGIN_STORY}}
How customers use it: {{SERVING_IDEAS}}
Output:
1. A 2–3 sentence opener on taste and texture, using specific sensory words from our tasting notes rather than "delicious" or "tasty".
2. A short origin paragraph using only the facts given.
3. "Try it": three serving or pairing ideas.
4. "Good to know": allergens in bold exactly as on the label, then storage, shelf life after opening, and net weight.
5. A one-line gift or occasion angle, if it fits.
Rules: copy allergens and ingredients exactly. Never summarize them as "free from" anything unless the label says so. No health or nutrition claims ("boosts immunity", "superfood", "detox", "good for your gut", "high in protein") unless I've confirmed the claim is permitted on the label in my market. Don't call it "artisan", "small-batch", "organic" or "local" unless the origin facts say so. Use only the facts I've given: no invented specs, certifications, ratings, rankings or quotes. List anything missing under "Facts to confirm".
Fill in: PRODUCT_NAME LABEL_TEXT TASTE_NOTES ORIGIN_STORY SERVING_IDEAS
Type or paste the label text exactly. The prompt copies allergens word for word and won't add nutrition or health claims.
Best for: Etsy and maker shops selling handmade or made-to-order pieces
You are a copywriter for independent makers. Write a product description for a handmade item that makes the craft visible and sets honest expectations.
Item: {{ITEM}}
Materials and how I make it (steps, tools, time taken): {{MATERIALS_AND_PROCESS}}
About me and the studio (facts I'm happy to share): {{MAKER_STORY}}
How each piece varies (glaze, grain, dye lots, sizes): {{VARIATION_NOTE}}
Made to order or ready to ship, and the lead time: {{LEAD_TIME}}
Write:
1. A two-sentence opening that puts the buyer in the moment of using or gifting the piece.
2. "How it's made": 3–4 sentences built from my process, with one concrete detail a factory product wouldn't have (such as the number of firings or a hand-stitched edge).
3. "Every piece is different": a warm, clear note on variation so buyers aren't surprised.
4. "Details": dimensions, materials, care and lead time.
5. A short personalization or gift-note line, if I offer one.
Rules: "handmade" covers only what I actually make by hand. If parts are bought in (findings, blanks, printed fabric), say so plainly. Don't claim "one of a kind" for a design I repeat. Use only the facts I've given: no invented specs, certifications, ratings, rankings or quotes. List anything missing under "Facts to confirm". Keep it under 220 words.
Fill in: ITEM MATERIALS_AND_PROCESS MAKER_STORY VARIATION_NOTE LEAD_TIME
Describe your real process step by step. The variation note stops buyers being surprised by differences between pieces.
Best for: Templates, presets, printables, fonts and courses
You are a copywriter for digital downloads (templates, presets, printables, courses, fonts). Write a sales description for {{PRODUCT_NAME}} that answers "what exactly do I get?" before anyone has to ask.
Contents (files, pages, lessons, counts): {{CONTENTS}}
File formats and what they open in: {{FORMAT_AND_COMPATIBILITY}}
What the buyer is trying to achieve: {{BUYER_GOAL}}
License and refund terms: {{LICENSE_TERMS}}
Output:
1. A one-sentence promise tied to the buyer's goal that doesn't promise an outcome (no income, follower or grade promises).
2. "What's inside": a checklist with the exact counts from my contents list.
3. "You'll need": the software, app or account required, and what it does NOT work with.
4. "How it works": purchase, download, first use, in three steps.
5. The license in plain English: personal vs commercial use, and what buyers can't do (resell, share, redistribute).
6. A three-question FAQ on the doubts digital buyers have: file access, editing and refunds.
Use only the facts I've given: no invented specs, certifications, ratings, rankings or quotes. List anything missing under "Facts to confirm". If my contents and license terms contradict each other (for example "commercial use" alongside fonts that restrict it), flag the conflict.
Fill in: PRODUCT_NAME CONTENTS FORMAT_AND_COMPATIBILITY BUYER_GOAL LICENSE_TERMS
List the contents with exact counts and the license terms you really offer. The prompt flags contradictions between them.
Best for: Resellers whose pages carry the same supplier text as every other store
You are an e-commerce editor. Many stores sell this product with the same manufacturer text, which gives shoppers and search engines no reason to choose my page. Rewrite it as original copy.
Manufacturer or supplier description: {{SUPPLIER_TEXT}}
Our brand voice: {{BRAND_VOICE}}
Our customer and why they buy from us: {{CUSTOMER}}
What we add that other sellers of this product don't (bundle, setup help, warranty handling, local stock, expert advice): {{DIFFERENTIATOR}}
Do this:
1. Extract every factual claim from the supplier text into a list and mark each one "keep", "needs checking" (vague, superlative or unverifiable, such as "world's best" or "eco") or "drop".
2. Write a new 120–180 word description from the "keep" facts only, in our voice, organized around our customer's use rather than the manufacturer's feature order.
3. Add a "Why buy from us" line that uses only our differentiator.
4. Check the result: no run of 8 or more words copied from the supplier text. List any phrases you had to keep, such as model names and technical terms.
Don't add features the supplier didn't state. If the supplier text contains claims that look regulated (health, safety, environmental), list them for me to verify before they go live.
Fill in: SUPPLIER_TEXT BRAND_VOICE CUSTOMER DIFFERENTIATOR
Paste the supplier text in full. Your differentiator is what makes the rewritten page worth choosing over other stores.
Amazon, Etsy, eBay & Google Shopping Listings
Every marketplace has its own format and its own rules. These prompts write Amazon titles that follow the January 2025 title requirements, five bullets, backend search terms and A+ modules, plus a compliance audit before you publish. They also cover Etsy titles and 13 tags, eBay titles with honest condition notes, Google Shopping feed titles, and one description adapted for four channels without the facts drifting.
Best for: New Amazon listings or titles flagged for policy issues
You are an Amazon listing specialist. Write the product title for this listing.
Product facts (type, key feature, material, size, color, quantity, model number): {{PRODUCT_FACTS}}
Brand name exactly as registered: {{BRAND}}
Amazon category or product type: {{CATEGORY_TEMPLATE}}
Keywords from my research, in priority order: {{TOP_KEYWORDS}}
Title character limit for my category (200 including spaces is the common cap; I'll confirm in my style guide): {{CHAR_LIMIT}}
Follow Amazon's title requirements as introduced in January 2025 (I'll check the current version):
- Start with the brand, then the product type and the most important differentiator.
- Stay within the character limit, including spaces.
- Don't use ! $ ? _ { } ^ ¬ ¦ unless they are part of the brand name.
- Don't use any word more than twice (articles, prepositions and conjunctions excepted).
- No promotional words (sale, free shipping, best seller, hot item), no subjective claims ("best", "amazing") and no ALL CAPS words except the brand or acronyms.
- Write numbers as numerals ("2 Pack", not "Two Pack").
Output three title options, each with its character count and the keywords it includes. Then tell me which option you'd use and why, show the first ~80 characters as they may appear on mobile, and list the keywords you left out that belong in the bullets or backend search terms. Use only facts I've given.
Fill in: PRODUCT_FACTS BRAND CATEGORY_TEMPLATE TOP_KEYWORDS CHAR_LIMIT
Check your category's title limit in Seller Central first. Use keywords from your own research tool, not guesses.
Best for: Writing or rewriting the 'About this item' section
You are an Amazon copywriter. Write the five "About this item" bullet points for this product.
Product facts and specs: {{PRODUCT_FACTS}}
Questions and doubts buyers have (from reviews, Q&A, returns, or competitor reviews I've read): {{CUSTOMER_QUESTIONS}}
Keywords to work in naturally: {{KEYWORDS}}
What's in the box: {{WHATS_INCLUDED}}
Character limit per bullet for my category: {{BULLET_LIMIT}}
Rules:
- Start each bullet with a short label (2–4 words, Title Case, not ALL CAPS), then a colon, then one concrete benefit backed by a fact.
- Order by buyer priority: the biggest reason to buy first, compatibility or sizing next, what's in the box last.
- One idea per bullet, within the limit. No emojis, no ™ ® © symbols and no decorative special characters.
- No pricing, shipping, promotional or company information, and no "satisfaction guaranteed" or refund promises. Warranty details go in the warranty field if my category has one.
- No claims like "eco-friendly", "antibacterial", "BPA-free" or "FDA approved" unless I've supplied proof. Some of these need certification, and some count as pesticide or health claims.
- Write for a person first; keywords must read naturally.
Output: the five bullets with character counts, a map showing which customer question each bullet answers, the questions left unanswered, and a "Facts to confirm" list.
Fill in: PRODUCT_FACTS CUSTOMER_QUESTIONS KEYWORDS WHATS_INCLUDED BULLET_LIMIT
Collect buyer questions from your reviews, Q&A, returns and competitors' reviews first. Each bullet should answer one of them.
Example output · Claude (example written by The Prompt Index)
Sample inputs: PRODUCT_FACTS = 24 oz (710 ml) double-wall vacuum-insulated food jar; 18/8 stainless steel interior; manufacturer test: hot up to 8 h, cold up to 12 h, starting with 95°C water, lid closed; mouth 3.1 in (79 mm); 6.1 in tall x 3.7 in wide (155 x 94 mm); screw lid with silicone gasket, manufacturer spec says "leakproof when closed with gasket fitted"; body hand wash, lid top-rack dishwasher safe. CUSTOMER_QUESTIONS = Does it really keep soup hot until lunch? Does it leak in a bag? Is it easy to clean? Will it fit my lunch bag? Is it safe for kids? Can it go in the microwave? KEYWORDS = insulated food jar, soup thermos, hot food container, leakproof lunch jar, wide mouth. WHATS_INCLUDED = jar, lid, folding stainless spoon stored in lid, spare gasket. BULLET_LIMIT = 255.
1. The five bullets
1) Stays Hot Until Lunch: Double-wall vacuum insulation keeps food hot for up to 8 hours and cold for up to 12 hours in the manufacturer's tests (filled with 95°C water, lid closed), so soup packed before work is still warm at midday. [231 characters]
2) Wide Mouth, Compact Size: Holds 24 oz (710 ml) in a jar 6.1 in tall and 3.7 in wide (155 x 94 mm), and the 3.1 in (79 mm) opening takes a full-size spoon, so you can eat straight from the jar. [192 characters]
3) Leakproof When Closed: The screw lid seals against a silicone gasket and is leakproof when closed with the gasket fitted, per the manufacturer's specification, so the jar can travel on its side in a lunch bag. [209 characters]
4) Easy to Clean: Hand wash the 18/8 stainless steel body; the lid can go on the dishwasher's top rack. A spare silicone gasket is included, so you can swap it when the first one wears. [182 characters]
5) What's in the Box: 1 insulated food jar (24 oz / 710 ml), 1 screw lid, 1 folding stainless steel spoon that stores inside the lid, and 1 spare silicone gasket. [159 characters]
2. Which question each bullet answers
- Bullet 1: "Does it really keep soup hot until lunch?" (keywords: insulated food jar, hot food container)
- Bullet 2: "Will it fit my lunch bag?" plus capacity, with exact dimensions so buyers can check (keyword: wide mouth)
- Bullet 3: "Does it leak in a bag?" (keyword: leakproof)
- Bullet 4: "Is it easy to clean?"
- Bullet 5: what's in the box
3. Questions left unanswered
- "Is it safe for kids?" I didn't target children or use "kids" as a keyword. A product designed or marketed mainly for children 12 and under counts as a children's product under US CPSIA rules, which require third-party testing and a Children's Product Certificate. Add the "kids" angle only if you hold one.
- "Can it go in the microwave?" Your facts don't say. Stainless steel jars normally can't, but confirm with the manufacturer and then state it plainly, because it is a common return reason for this product type.
- Keyword "soup thermos" isn't used: Thermos is a brand name and a trademark in many countries, and other brands' names don't belong in your listing. Use "insulated soup jar" in the backend terms instead.
4. Facts to confirm
- Is the 8 h / 12 h test result printed on your packaging or in a manufacturer document you can produce if asked?
- Get the leakproof specification in writing from the manufacturer.
- Bullet 1 says "still warm at midday", which assumes food is packed about 4–5 hours before lunch. Cut that clause if you want to make no claim beyond the test figures.
- I left out "BPA-free" because no test report was supplied. If the lid supplier gives you one, it can go in bullet 4.
Best for: Filling the generic keyword field with terms the listing doesn't already cover
You are an Amazon SEO specialist. Build the backend search terms (generic keyword) field for my listing.
My current title and bullets: {{TITLE_AND_BULLETS}}
My keyword research (terms, plus volume if I have it from a tool): {{KEYWORD_LIST}}
Byte limit for the field in my marketplace (commonly under 250 bytes; I'll check Seller Central): {{BYTE_LIMIT}}
Do this:
1. Remove every word that already appears in the title or bullets. They're already indexed, so repeating them wastes space.
2. Remove brand names (mine and competitors'), ASINs, temporary words ("new", "on sale"), subjective claims ("best", "cheapest") and offensive terms, none of which belong in this field.
3. Keep synonyms, alternate spellings, abbreviations, use cases and other-language terms common in my marketplace. Don't add both singular and plural forms or common misspellings, which Amazon generally matches already.
4. Output the final string: lowercase, single spaces, no commas, no repeated words, within the byte limit.
5. Show the byte count, counting accented or multi-byte characters such as é as 2 bytes.
6. List the keywords that didn't fit, ranked, so I can rotate them in when testing.
Don't invent search volumes. If I gave none, rank by relevance and say that's what you did.
Fill in: TITLE_AND_BULLETS KEYWORD_LIST BYTE_LIMIT
Paste your live title and bullets so the prompt can strip duplicates. Check the byte limit for your marketplace.
Best for: Brand-registered sellers planning A+ (Enhanced Brand Content)
You are an Amazon A+ Content strategist. Plan the A+ section for this product and write its copy.
Product facts: {{PRODUCT_FACTS}}
Brand story (facts only): {{BRAND_STORY}}
Questions the title and bullets don't fully answer: {{CUSTOMER_QUESTIONS}}
Images and assets I have or can shoot: {{AVAILABLE_ASSETS}}
Other products in my range (for a comparison chart): {{PRODUCT_RANGE}}
Output:
1. A plan of 5–7 modules in order (for example header banner, three-image feature row, comparison chart, how to use, brand story) and the buyer question each one answers.
2. For each module: headline, body copy that fits typical module text lengths, and an image brief (what the image shows, plus any text overlay of 6 words or fewer).
3. A comparison chart across my range using only facts I gave, with a "best for" row phrased as a use case, not a ranking.
4. Alt text for every image: accurate and descriptive for shoppers using screen readers, with keywords only where they describe the image.
Avoid what A+ reviews commonly reject: pricing or promotions, shipping details, customer reviews or quotes, "best-selling" or "top-rated" claims, warranty or guarantee language, time-sensitive words, and links or contact details. Use only the facts I've given: no invented specs, certifications, ratings, rankings or quotes. List anything missing under "Facts to confirm".
Fill in: PRODUCT_FACTS BRAND_STORY CUSTOMER_QUESTIONS AVAILABLE_ASSETS PRODUCT_RANGE
List the images you actually have or can shoot. The module plan is built around them.
Best for: Catching suppression risks and unsupported claims before a listing goes live
You are a marketplace compliance reviewer. Audit this Amazon listing before it goes live, so it isn't suppressed, edited by Amazon or reported by a competitor.
Full listing (title, bullets, description, backend terms, A+ text): {{LISTING_TEXT}}
Product category: {{PRODUCT_CATEGORY}}
Certifications, test reports and registrations I actually hold: {{CERTIFICATIONS_HELD}}
Check the following and report in a table (location | exact text | issue | risk level | suggested fix):
1. Title rules: length, banned characters, words used more than twice, promotional words.
2. Prohibited or restricted claims: medical or disease claims; pesticide-type claims ("antibacterial", "antimicrobial", "kills germs"); environmental claims ("eco-friendly", "biodegradable", "compostable", "non-toxic") without matching proof; "FDA approved" or "FDA registered" wording; and "guaranteed" or refund language.
3. Ranking and popularity claims: "best seller", "#1", "top rated", "Amazon's Choice".
4. Competitor brand names or trademarks used as keywords.
5. Pricing, shipping details, contact details or URLs anywhere in the copy.
6. Any claim the certifications list doesn't support.
Then give the cleaned listing with every fix applied, keeping the selling points that survive. Note where a rule varies by category or marketplace so I can confirm it in Seller Central. This is a risk check, not legal advice.
Fill in: LISTING_TEXT PRODUCT_CATEGORY CERTIFICATIONS_HELD
Paste every text field of the listing. List only certifications and test reports you actually hold.
Best for: Etsy sellers who want readable titles and tags that cover different searches
You are an Etsy SEO coach. Write the title, tags and attribute suggestions for this listing.
The item, in my words: {{ITEM_DESCRIPTION}}
Phrases I think buyers search (from Etsy search suggestions or my shop stats): {{BUYER_SEARCHES}}
Occasions, recipients or rooms it suits: {{OCCASIONS}}
Materials, style, color and techniques: {{MATERIALS_AND_STYLE}}
Rules:
- Title: up to 140 characters, but say clearly what the item is within the first ~40 characters. Etsy's own guidance favors short, readable titles over keyword lists, so no repeated phrases and no comma-separated strings of keywords.
- Tags: exactly 13, each 20 characters or fewer, using multi-word phrases rather than single words, with no duplicates, and covering different angles: what it is, style, material, occasion, recipient and use.
- Don't spend tags on things the attributes (color, material, occasion, holiday) already cover. List those attributes separately.
- Don't use other brands, trademarked characters, or phrases like "Disney-style" in tags or the title.
Output: two title options with character counts; the 13 tags as a numbered list with character counts; the attributes to set; and the first two lines of the description, which show in search previews, using only the facts I gave.
Fill in: ITEM_DESCRIPTION BUYER_SEARCHES OCCASIONS MATERIALS_AND_STYLE
Pull search phrases from Etsy's search bar suggestions and your shop stats. Set attributes too, so tags aren't wasted on them.
Best for: Used, refurbished and collectible items where honesty about condition sells
You are an eBay power-seller coach. Write the listing for this item so buyers trust it and search finds it.
Item details (brand, model, MPN, size, color, year, compatibility): {{ITEM_DETAILS}}
Condition (new, open box, refurbished and by whom, used): {{CONDITION}}
Every flaw, mark, missing part or fault I know about: {{FLAWS}}
What's included and what isn't: {{INCLUDED_ITEMS}}
Typical buyer (collector, bargain hunter, trade, parts): {{BUYER_TYPE}}
Output:
1. Title: up to 80 characters, built from what buyers type (brand, model, key spec, size, and a condition word if it helps). No "L@@K", "wow" or keyword spam, and no brand names other than the item's own.
2. Item specifics: a two-column list of the fields eBay asks for in this category, filled from my details, with "unknown, please check" where I gave no value.
3. Condition description: honest and specific. Every flaw from my list, where it is and how visible it is, plus a line suggesting which photos show it.
4. Description (80–150 words): what it is, what's included, what's not, and compatibility.
5. A "Before you buy" line covering returns and postage, only as I state them.
Never describe the condition more favorably ("like new", "mint") than my flaws list supports, and don't write "refurbished" unless I say who refurbished it.
Fill in: ITEM_DETAILS CONDITION FLAWS INCLUDED_ITEMS BUYER_TYPE
List every flaw you know about. Honest condition notes prevent 'not as described' returns and disputes.
Best for: Merchant Center feeds with weak or truncated product titles
You are a product feed specialist for Google Merchant Center. Rewrite feed titles and descriptions so products match more relevant shopping searches.
Feed rows with headers (id, title, description, brand, color, size, material, gtin, product_type): {{FEED_ROWS}}
What these products are: {{PRODUCT_TYPE}}
Search terms from my Shopping search terms report, if I have one: {{SEARCH_TERMS}}
Rules:
- Title: max 150 characters, with the most important words in the first ~70, because Shopping often truncates there. Use an attribute order that suits the product type: apparel is brand + gender + product type + color, size, material; hard goods are brand + product + key spec + model or size.
- Every word in a title must come from that row's data. Don't add attributes the row doesn't have.
- No promotional text ("free shipping", "sale", "% off"), no ALL CAPS and no decorative symbols.
- Description: 500–1,000 characters, product facts first, with no comparisons to other stores and no promotional text.
Output a table with the columns id | new title | character count | new description | attributes missing from the row (for example, no material listed). Keep the same ids and row order so I can paste it back. Then list three patterns you changed across the feed, so I can apply them to products you didn't see.
Fill in: FEED_ROWS PRODUCT_TYPE SEARCH_TERMS
Paste 10–30 feed rows at a time with their headers. The output keeps your ids and order so it pastes back into the sheet.
Best for: Selling one product on Shopify, Amazon, Etsy and eBay without facts drifting
You are a multichannel e-commerce copywriter. I sell the same product on several channels and need each version to fit that channel's format and shopper, without the facts drifting between them.
Master description from my own store: {{MASTER_DESCRIPTION}}
Verified product facts (the source of truth): {{PRODUCT_FACTS}}
Channels I sell on (for example Shopify, Amazon, Etsy, eBay): {{CHANNELS}}
For each channel, produce:
1. The adapted copy in that channel's usual structure: a story-led Shopify page in HTML; an Amazon title, five bullets and a plain description; an Etsy readable title, first two lines and details; an eBay title of up to 80 characters and a specifics-led description. For any other channel, tell me the structure you're assuming.
2. Tone notes: what changed and why (Amazon shoppers scan specs; Etsy shoppers care about the maker and the gift angle).
3. A consistency table: each key fact (dimensions, materials, quantity, warranty, care) and its wording on every channel.
Rules: facts stay identical across channels; only emphasis and format change. Strip anything a channel restricts, such as links, promotions and contact details on marketplaces. Use only the facts I've given: no invented specs, certifications, ratings, rankings or quotes. List anything missing under "Facts to confirm". Tell me which channel limits to confirm before publishing.
Fill in: MASTER_DESCRIPTION PRODUCT_FACTS CHANNELS
Give the verified facts separately from the copy. The consistency table shows any fact that differs between channels.
Product & Collection Page SEO
Product pages rank and get clicked when the title, snippet and on-page answers match what people search. These prompts write SEO titles and meta descriptions, turn a bare collection grid into a page with buying advice, map keywords to pages so they don't compete, build Product schema from real data only, and write a fair 'X vs Y' buyer's guide. They also write alt text that helps blind shoppers and a Q&A block that AI shopping assistants can quote.
Best for: Product pages with default or missing SEO titles and meta descriptions
You are an e-commerce SEO specialist. Write the SEO title and meta description for a product page.
Product: {{PRODUCT_NAME}}
Main keyword from my research: {{PRIMARY_KEYWORD}}
Facts that make this product worth clicking (material, size, use, compatibility): {{KEY_FACTS}}
Store name: {{STORE_NAME}}
True offer facts I can mention (free delivery threshold, returns window, where it's made): {{OFFER_FACTS}}
Write:
1. Three SEO titles of 50–60 characters including spaces, with the main keyword near the start and the store name at the end only if it fits.
2. Three meta descriptions of 140–155 characters that say what the product is, give one concrete reason to choose it and one true offer fact, and end with a plain call to action ("Shop sizes and colors").
3. A character count for each.
4. The title and description pairing you recommend, and why it fits the search intent (someone comparing vs someone ready to buy).
5. A suggested URL handle: lowercase, hyphens, no stop words, no dates.
Rules: no keyword stuffing (the keyword appears once per title and at most once per description), no ALL CAPS, no "best" or "cheapest" unless I supply evidence, and no offers I didn't list. Search engines may rewrite titles and snippets, so aim for accurate rather than clever.
Fill in: PRODUCT_NAME PRIMARY_KEYWORD KEY_FACTS STORE_NAME OFFER_FACTS
Bring the main keyword from your own research or Search Console. Only list offers you actually run.
Best for: Category and collection pages that are just a product grid
You are an SEO copywriter for e-commerce category pages. Write the copy for my {{COLLECTION_NAME}} collection page, which is currently just a product grid.
Products in this collection (names, key differences, price bands): {{PRODUCTS_IN_COLLECTION}}
Questions shoppers ask when choosing in this category: {{SHOPPER_QUESTIONS}}
Keywords to cover (main keyword plus variations): {{TARGET_KEYWORDS}}
Related collections I can link to: {{RELATED_COLLECTIONS}}
Write:
1. A top-of-page intro (50–80 words) above the grid: what's in the collection and how to choose, kept short so the products stay visible.
2. A "How to choose" block (200–300 words) for below the grid, with subheads for the 2–4 decisions shoppers face (size, material, use case), naming specific products from my list as examples.
3. A 4–5 question FAQ with short answers based on my product data.
4. Internal links: anchor text and target collection for 3–5 links.
5. An SEO title (60 characters or fewer) and meta description (155 characters or fewer).
Rules: write for a shopper deciding, not for a keyword counter. Use the main keyword in the H1, in the intro and once in the buying advice; use variations naturally elsewhere. No "best" rankings unless they rest on a stated criterion ("best for small spaces: X, because it's 30 cm deep").
Fill in: COLLECTION_NAME PRODUCTS_IN_COLLECTION SHOPPER_QUESTIONS TARGET_KEYWORDS RELATED_COLLECTIONS
List the products in the collection with their real differences. The buying advice names them as examples.
Best for: Stopping product, collection and blog pages competing for the same searches
You are an e-commerce SEO strategist. Map my keyword research to the pages on my store so each page targets one clear intent and my pages don't compete with each other.
Keyword export (keyword, monthly volume, difficulty if available) from my SEO tool or Search Console: {{KEYWORD_EXPORT}}
My current pages (URL, page type: product / collection / blog / guide, current title): {{PAGE_LIST}}
What the store sells and to whom: {{STORE_FOCUS}}
Do this:
1. Group the keywords by search intent: buying a specific product, browsing a category, comparing options, or learning.
2. Assign each group to the single best existing page: product pages for model-specific terms, collection pages for category terms, and guides for "how to" and "vs" terms.
3. Flag cannibalization, where two or more of my pages would target the same group, and recommend which page should own it.
4. Flag gaps: keyword groups with no suitable page, and the page type to create for each.
5. For each page, give the primary keyword, 2–4 secondary terms and a suggested title.
Output a table I can paste into a spreadsheet: page | type | primary keyword | secondary keywords | combined volume | action (keep / retitle / create / merge).
Use only the volumes in my export. Don't estimate missing volumes; mark them "no data".
Fill in: KEYWORD_EXPORT PAGE_LIST STORE_FOCUS
Export keywords from your SEO tool or Search Console and list your pages with their types. Gaps become your page to-do list.
Best for: Adding Product and Offer structured data without risking a manual action
You are a technical SEO developer. Write Product structured data (JSON-LD, schema.org) for one product page, using only the data I give you.
Product data (name, description, SKU, GTIN/MPN, brand, image URLs, price, currency, availability, condition, variants): {{PRODUCT_DATA}}
Review and rating data shown visibly on the page, if any (average, count): {{REVIEW_DATA}}
Page URL: {{PAGE_URL}}
Produce:
1. A JSON-LD block with Product and Offer (or ProductGroup with hasVariant if I listed variants). Include shippingDetails and hasMerchantReturnPolicy only if I supplied shipping and returns facts.
2. Include aggregateRating ONLY if I gave real review data that is visible on the page. If there's no review data, leave it out entirely: ratings that are invented or not shown on the page break Google's structured data policies.
3. A field checklist of the required and recommended properties for product rich results, marked present or missing, with what I need to supply.
4. Notes on which values must match the visible page (price, availability) and how to keep them in sync when the price changes.
Rules: valid JSON (double quotes, no trailing commas, no comments inside the block). Use ISO 4217 currency codes and schema.org availability URLs such as https://schema.org/InStock. Don't invent GTINs, prices or review counts. Remind me to run Google's Rich Results Test.
Fill in: PRODUCT_DATA REVIEW_DATA PAGE_URL
Leave the review data empty if ratings aren't visible on the page. Test the output in Google's Rich Results Test before publishing.
Best for: Accessible, descriptive alt text for every image in a gallery
You are an accessibility and image-SEO specialist for online stores. Write alt text and file names for every image in a product gallery.
Product: {{PRODUCT_NAME}}
Images in gallery order, each with a short note on what it shows (angle, color, detail, lifestyle scene, size chart): {{IMAGE_LIST}}
Variants the images belong to (colors, finishes): {{VARIANTS}}
For each image, give:
1. Alt text of 8–20 words describing what a blind shopper needs from that image: color, angle, the detail shown, and scale cues. Start with the product, never with "Image of" or "Picture of".
2. A file name: lowercase, hyphenated and descriptive (product-color-view.jpg), with no camera numbers.
3. A flag if the image contains text (size chart, infographic). That text must also appear in the page copy, because alt text shouldn't carry a whole chart.
Rules:
- Every alt text must be different. Six identical alt texts help no one.
- Mention the variant color only on images that show it.
- Use the product's search terms only where they naturally describe the image; never stuff keywords.
- Decorative images (background textures, spacers) get empty alt text; say which ones.
Output a table: position | file name | alt text | notes.
Fill in: PRODUCT_NAME IMAGE_LIST VARIANTS
List the images in gallery order with a few words on what each shows. Works for Shopify, WooCommerce and most store builders.
Best for: Making product pages answer the questions shoppers ask AI assistants
You are an e-commerce content strategist. AI shopping assistants such as Amazon's Rufus, ChatGPT and Google's AI Mode answer buyer questions using what's written on product pages, reviews and Q&A. Write a product Q&A block that gives them, and human shoppers, accurate answers to use.
Verified product facts and specs: {{PRODUCT_FACTS}}
Real questions buyers have asked (support emails, chat logs, marketplace Q&A, reviews): {{REAL_QUESTIONS}}
Shipping, returns and warranty policies: {{POLICIES}}
Produce:
1. 8–12 questions in the words buyers actually use, grouped into fit and compatibility, use and care, comparison with alternatives, and delivery and returns.
2. Answers of 1–3 sentences that start with a direct yes, no or number, then give the reason, using only the facts above.
3. An attribute gap list: questions you couldn't answer from my facts. These are also the gaps an AI assistant may fill with guesses or a competitor's data, so they're my to-do list.
4. The Q&A in plain Question / Answer pairs I can paste into the page.
Rules: don't promise outcomes, and state limits plainly ("Not dishwasher safe"). Use only the facts I've given: no invented specs, certifications, ratings, rankings or quotes. List anything missing under "Facts to confirm". Don't claim to know how assistants rank products.
Fill in: PRODUCT_FACTS REAL_QUESTIONS POLICIES
Use real questions from support, chat and marketplace Q&A. The gap list is your to-do list of missing product facts.
Best for: 'X vs Y' searches that happen before someone picks a product
You are a product expert writing a buyer's guide for an online store. Shoppers search "X vs Y" before they buy. Write a fair comparison guide that helps them choose and links to the products we sell.
Category: {{CATEGORY}}
The product types or materials being compared (for example cast iron vs stainless steel vs nonstick): {{PRODUCT_TYPES}}
Our products of each type, with facts: {{OUR_PRODUCTS}}
The customer situations we see most often: {{CUSTOMER_SITUATIONS}}
Write a 900–1,300 word guide:
1. The short answer first: which type suits which situation, in 3–4 bullets.
2. A comparison table across the 6–8 factors that matter (cost, care, durability, weight, compatibility), using widely accepted general characteristics of each type and marking anything that varies by brand.
3. One section per type: strengths, weaknesses, who it's for and who should skip it.
4. A "Situations" section: for each customer situation, the type you'd recommend and why.
5. Where natural, a link suggestion to our product of that type, with anchor text.
6. An SEO title (60 characters or fewer) and meta description (155 characters or fewer).
Rules: include honest weaknesses of the types we sell. Don't cite tests, statistics or health effects (such as coating safety) unless I give a source; mark any claim that needs one with [source needed].
Fill in: CATEGORY PRODUCT_TYPES OUR_PRODUCTS CUSTOMER_SITUATIONS
Name the product types or materials shoppers weigh up. Honest weaknesses of what you sell make the guide credible.
Bulk Catalog Workflows From a Spreadsheet
Most stores don't need one great description; they need three hundred good ones. Paste spreadsheet rows and these prompts clean the attribute data, write descriptions in batches, build a template for product families and fact-check every AI-written draft against its source row. There is also a localization prompt for selling into a new country, and one that uses your analytics to decide which pages to rewrite first.
Best for: Writing 10–25 product descriptions at a time from catalog data
You are a catalog copywriter working in batches. Write product descriptions for every row of spreadsheet data I paste, consistently and without inventing anything.
Column headers: {{COLUMN_HEADERS}}
Rows (10–25 at a time, tab- or comma-separated): {{ROWS}}
Brand voice card or style notes: {{BRAND_VOICE}}
Output format I need (for example Shopify CSV columns Handle, Body (HTML), SEO Title, SEO Description): {{OUTPUT_FORMAT}}
Process:
1. First, list any rows with missing or contradictory data (a blank material, sizes in two units that disagree) and don't write those rows. Don't fill gaps with guesses.
2. For every complete row, write a one-sentence hook, three benefit bullets drawn from that row's columns, and a details line: 80–120 words in total.
3. Vary the opening structure across rows so the products don't all start the same way. No two descriptions may share an opening phrase.
4. Return the batch in my output format, one row per product, keeping my identifier column (SKU or handle) unchanged and in the same order.
5. After the batch, list the rows I should check by hand, and why.
Every fact must come from its own row. Never move a feature from one product to another. If a value looks wrong (an 8 oz candle with a 450-hour burn time), flag it rather than writing around it.
Fill in: COLUMN_HEADERS ROWS BRAND_VOICE OUTPUT_FORMAT
Paste rows in batches of 10–25 with the header row. Rows with missing or odd data are flagged and skipped, not guessed.
Example output · Claude (example written by The Prompt Index)
Sample inputs: COLUMN_HEADERS = SKU | Name | Scent notes | Wax | Burn time (hrs) | Size | Wick | Vessel. ROWS =
CN-101 | Harbour Morning | sea salt, lemon peel, driftwood | coconut-soy blend | 45 | 8 oz / 227 g | cotton | amber glass jar, tin lid
CN-102 | Orchard Kitchen | baked apple, clove, brown sugar | coconut-soy blend | 45 | 8 oz / 227 g | cotton | amber glass jar, tin lid
CN-103 | Night Library | cedarwood, leather, black pepper | coconut-soy blend | 450 | 8 oz / 227 g | wood | matte black ceramic
CN-104 | Fig & Linen | green fig, white tea, cotton | (blank) | 60 | 12 oz / 340 g | cotton | clear glass tumbler
CN-105 | Ember Road | smoked vanilla, tobacco leaf, amber | coconut-soy blend | 70 | 14 oz / 397 g | 2 cotton wicks | amber glass jar, tin lid
BRAND_VOICE = warm, plain, a little wry; never "luxurious", "indulge" or "clean-burning". OUTPUT_FORMAT = plain text, one block per product headed by the SKU, same order.
1. Rows not written
- CN-103 Night Library: burn time is 450 hours, but the other 8 oz candles are 45 hours. It looks like a typo for 45. Please confirm before I write it.
- CN-104 Fig & Linen: the Wax column is blank. I won't guess the wax type, because buyers filter and choose on it.
2. Descriptions
CN-101 | Harbour Morning
Sea salt and lemon peel first, with driftwood underneath: this is the walk along the water before anyone else is up.
- A bright top of sea salt and lemon peel that keeps a room feeling awake
- A driftwood base that gives the scent some weight, so it isn't just citrus
- A rated burn time of 45 hours from one 8 oz candle
Details: 8 oz / 227 g. Coconut-soy blend wax, cotton wick. Amber glass jar with a tin lid. Burn time 45 hours.
CN-102 | Orchard Kitchen
If your kitchen has ever smelled of an apple pie that nobody has actually baked, you already know this one. Light it on the first evening that needs a sweater.
- Baked apple up front, rounded off with brown sugar
- Clove adds warmth without tipping into potpourri
- Rated for 45 hours of burning from one 8 oz candle
Details: 8 oz / 227 g. Coconut-soy blend wax, cotton wick. Amber glass jar with a tin lid. Burn time 45 hours.
CN-105 | Ember Road
Two cotton wicks and 14 oz of wax, built for long evenings that end later than planned.
- Smoked vanilla and tobacco leaf: sweet, but with smoke at the edges
- Amber at the base keeps it warm and steady once the room has filled
- Twin cotton wicks and a 70-hour rated burn time
Details: 14 oz / 397 g. Coconut-soy blend wax, 2 cotton wicks. Amber glass jar with a tin lid. Burn time 70 hours.
3. Check by hand
- All rows: the burn-time column doesn't say how it was tested (wick trimmed, burn length per session). Keep the word "rated" unless you can say how you measured it.
- All rows: no safety line is included because none was in the data. Your page and label should carry the candle safety warnings your market requires (in the UK and EU, scented candles may also need CLP hazard labeling for fragrance allergens). Add these from your label, not from me.
Best for: Messy supplier spreadsheets with inconsistent colors, units and sizes
You are a product data manager. Before any copy is written, clean the attribute data in my catalog so descriptions, filters and feeds all use the same values.
Raw product rows (with headers): {{RAW_ROWS}}
What these products are: {{PRODUCT_CATEGORY}}
Our preferred conventions, if any (units, color names, size labels, capitalization): {{ATTRIBUTE_RULES}}
Do this:
1. Profile each column: distinct values, blanks, and inconsistent forms (for example "Navy", "navy blue", "NVY", or "30cm", "30 cm", "11.8 in").
2. Propose a standard value list for each column that follows my conventions, or sensible defaults if I gave none, and state the defaults you chose.
3. Return the cleaned rows as a table with the same identifier column and order, plus a "changes" column describing what you changed in that row.
4. List anything you couldn't standardize safely, such as values that could mean two things, and ask me to decide.
5. Suggest attributes shoppers filter by in this category (material, fit, capacity, compatibility) that are missing, and which rows lack them.
Rules: standardize the format, never the fact. Don't convert units unless I ask, and if I do, keep the original value in brackets. Don't merge two colors or sizes because they look similar.
Fill in: RAW_ROWS PRODUCT_CATEGORY ATTRIBUTE_RULES
Run this before the bulk description prompt. Clean data gives you better copy, filters and feeds.
Best for: Many near-identical variants that each need a useful, distinct page
You are an e-commerce content designer. I sell many near-identical products (the same item in different sizes, scents, patterns or capacities). Design a description template that keeps each page distinct and useful without rewriting everything each time.
Product family: {{PRODUCT_FAMILY}}
Three sample products with their full details: {{SAMPLE_PRODUCTS}}
Attributes that change between products: {{VARYING_ATTRIBUTES}}
Brand voice: {{BRAND_VOICE}}
Produce:
1. A template with fixed blocks (family story, materials, care) and slot blocks marked [SLOT: attribute] for what changes.
2. For each slot, 3–5 sentence patterns so variants don't read identically, and the rule for choosing between them.
3. A "unique angle" slot: one sentence per product on the specific use or person that variant suits (the 2-cup size for one person, the 8-cup size for a household).
4. The three sample products written out with the template, to prove it works.
5. Guidance on when a variant deserves its own page and when it should be an option on one page, so you don't create thin, duplicate pages.
Keep the shared blocks short so the unique part of each page is a meaningful share of its text. Use only the facts I've given: no invented specs, certifications, ratings, rankings or quotes. List anything missing under "Facts to confirm".
Fill in: PRODUCT_FAMILY SAMPLE_PRODUCTS VARYING_ATTRIBUTES BRAND_VOICE
Give three real products from the family. The template is tested on them before you use it on the rest.
Best for: QA before publishing any AI-generated catalog copy
You are a meticulous product-content QA editor. AI-written descriptions sometimes add features, round numbers or borrow details from another product in the batch. Audit this batch before it goes live.
Source data (the truth), one row per product with an identifier: {{SOURCE_ROWS}}
Draft descriptions, each labeled with the same identifier: {{DRAFT_DESCRIPTIONS}}
Claims we must never make without proof (our list, for example "waterproof", "organic", "non-toxic", "clinically tested"): {{RESTRICTED_CLAIMS}}
For each product, check:
1. Every number, material, dimension, capacity and compatibility statement in the draft against its own source row.
2. Features in the draft that aren't in the source, whether invented or carried over from a neighboring product.
3. Restricted or regulated claims, and superlatives ("best", "strongest", "longest-lasting").
4. The wrong product name, variant or color.
5. Near-duplicate wording shared with other drafts in the batch.
Output a table: identifier | issue type | exact draft text | what the source says | fix. Then give a pass/fail list by identifier and the corrected text for every failed description. If the source itself looks wrong, mark it "source check" rather than changing the copy to match. Be strict: if you can't find a claim in the source, it fails.
Fill in: SOURCE_ROWS DRAFT_DESCRIPTIONS RESTRICTED_CLAIMS
Label each draft with the same SKU or handle as its source row. Anything the checker can't trace to the source fails.
Best for: Expanding a store or marketplace listing into a new country
You are an e-commerce localization editor. Adapt my product copy for a new market. This is localization, not just translation: units, sizing, spelling, references and legal wording all need to fit.
Source copy: {{SOURCE_COPY}}
Selling from (market and language): {{SOURCE_MARKET}}
Selling into (country and language): {{TARGET_MARKET}}
Verified product facts, including any market-specific versions (plug type, voltage, size standard): {{PRODUCT_FACTS}}
Deliver:
1. The localized copy, written as a native would write it, not translated word for word.
2. Units and sizes converted, with the original in brackets where buyers may compare (for example "28 cm (11 in)"). Map clothing and shoe sizes only with a conversion I supplied; otherwise flag it for me to supply.
3. Changed references: idioms, holidays, seasons (including the southern hemisphere), currency mentions and spelling.
4. Claims or words that may need different treatment in the target market: regulated claims, required labeling, words with a different meaning, and anything that triggers a local rule I should check.
5. SEO: local search phrases worth researching (don't invent volumes), plus a localized SEO title and meta description.
Rules: don't change any fact. If a product fact differs between markets (voltage, plug), use the target-market fact, or ask me for it.
Fill in: SOURCE_COPY SOURCE_MARKET TARGET_MARKET PRODUCT_FACTS
Include any market-specific product facts (plug type, voltage, size standard). Have a native speaker review before launch.
Best for: Deciding where copy work will move revenue first
You are a conversion-focused e-commerce editor. Decide which product pages to rewrite first, then fix the top one.
Page performance export for the last 90 days (product, sessions, add-to-cart rate, conversion rate, return rate, top return reasons): {{PAGE_DATA}}
Current copy of the top-priority page (paste after step 2 if you prefer): {{CURRENT_COPY}}
What has changed since these pages were written (new photos, new reviews, price, stock, product updates): {{CHANGES_SINCE_LAUNCH}}
Do this:
1. Sort each product into a pattern: high traffic but low add-to-cart (the copy or photos may be the problem), good add-to-cart but low conversion (price, shipping or trust at checkout), or a high return rate (an expectation gap).
2. Rank the top five pages to fix and say what the data suggests is wrong with each, labeled as a hypothesis, not a conclusion.
3. For the first page, diagnose the copy against its numbers: missing information, buried key facts, an unclear "who it's for", or a mismatch with the return reasons.
4. Rewrite it, and list every change so I can A/B test it or compare before and after.
5. Suggest how long to wait and what to measure before judging the change.
Use only my numbers. Don't quote industry benchmark conversion rates as facts; if you mention a typical range, label it an assumption to check.
Fill in: PAGE_DATA CURRENT_COPY CHANGES_SINCE_LAUNCH
Export 90 days of product-level analytics, with return reasons if you have them. Treat the diagnosis as a hypothesis to test.
Variants, Size Guides & Product Details
The details under the main description decide whether the order comes back. These prompts name color and size variants so shoppers can tell them apart, turn your measurements into a size guide, write care and what's-in-the-box copy, and build a good-better-best comparison across your range. A bundle prompt calculates the real saving instead of an inflated 'worth' figure.
Best for: Colors, finishes and scents that shoppers struggle to tell apart
You are a merchandising copywriter. Write variant names and short variant descriptions for {{PRODUCT_NAME}} so shoppers can tell the options apart without opening every photo.
Variants (SKU, current name, color, finish, size or scent, and any spec differences): {{VARIANT_LIST}}
What actually differs between them beyond appearance (capacity, weight, price, materials): {{WHAT_DIFFERS}}
Produce:
1. Variant names that are clear and evocative but always anchored to the real color or finish ("Sage Green", not just "Meadow"). Keep a standard color word in every name so filters and color searches still work.
2. A one-line description per variant (max 20 words) saying what it looks like in real light and any functional difference.
3. A "How to choose" line for the product page that explains the main trade-off between variants in one or two sentences (for example size vs portability).
4. Swatch label text, max 15 characters each.
5. A note on any variants so similar that shoppers will struggle (two near-identical grays), and how to describe the difference honestly, or whether to drop one.
Rules: describe colors truthfully. Screens vary, so add a line telling shoppers how to judge the color, such as a natural-light photo or a swatch option. Use only the facts I've given: no invented specs, certifications, ratings, rankings or quotes. List anything missing under "Facts to confirm".
Fill in: PRODUCT_NAME VARIANT_LIST WHAT_DIFFERS
List every variant with its SKU and real differences. Keep a plain color word in each name so filters still work.
Best for: Clothing and footwear stores fighting wrong-size returns
You are a fit specialist for an online clothing or footwear store. Turn my measurement chart into a size guide shoppers can actually use, to cut wrong-size returns.
Product type: {{PRODUCT_TYPE}}
Measurements for each size, exactly as measured: {{MEASUREMENT_CHART}}
Fit feedback from customers or returns ("runs small in the shoulders" and so on): {{FIT_FEEDBACK}}
Units to show (cm, inches or both): {{UNITS}}
Deliver:
1. A size table in the units requested, clearly labeled as body measurements or garment measurements. Never mix the two in one table, and if my chart is garment measurements, say so.
2. "How to measure yourself": 3–5 short numbered steps for the measurements that matter for this product, written so someone can do it alone with a tape measure.
3. "Between sizes?": a clear rule based on the fit feedback (for example "size up if you like room through the shoulders").
4. A fit summary line for the product page, max 25 words.
5. Conversions to other sizing systems (UK, US, EU) ONLY if I supplied them; otherwise list them as a gap. Brands size differently, so generic conversion charts cause returns.
Check my chart for errors before using it: sizes that don't increase steadily, or values that look mistyped. List them first.
Fill in: PRODUCT_TYPE MEASUREMENT_CHART FIT_FEEDBACK UNITS
Paste your own measured chart and say whether it's body or garment measurements. Conversions are only added if you supply them.
Best for: Product page tabs that prevent misuse, support tickets and returns
You are a product-support copywriter. Write the care, use and "what's in the box" sections for {{PRODUCT_NAME}} so customers use it correctly and don't contact support or return it for avoidable reasons.
Manufacturer care and use instructions (pasted exactly): {{MANUFACTURER_CARE}}
Contents of the box or package: {{BOX_CONTENTS}}
Mistakes customers make, from support tickets or returns: {{COMMON_MISTAKES}}
Write:
1. "What's in the box": a checklist with quantities, plus a "Not included" line for things people assume come with it (batteries, plug adapter, refills).
2. "Before first use": up to four steps.
3. "Care": short do and don't lines in plain English, rewritten from the manufacturer's wording without changing its meaning. Temperatures, cycles and products stay exactly as stated.
4. "Avoid these mistakes": one line per common mistake and how to prevent it.
5. Safety warnings from the manufacturer text, copied without softening.
Rules: never add, drop or relax a care or safety instruction. If the manufacturer text is ambiguous ("wash cool"), keep it and flag the ambiguity for me. Format it for a product page tab or accordion: short headings, bullets, and no paragraph longer than two lines.
Fill in: PRODUCT_NAME MANUFACTURER_CARE BOX_CONTENTS COMMON_MISTAKES
Paste the manufacturer's care and safety text exactly. Pull common mistakes from support tickets or return notes.
Best for: Product ranges where shoppers stall between models at different prices
You are a merchandising strategist. I sell several versions of one product at different prices, and shoppers stall because they can't see the difference. Build a comparison that helps each shopper pick the right one.
Every model in the range, with full specs: {{RANGE_DETAILS}}
The different buyers who shop this range and what they need: {{BUYER_TYPES}}
Prices: {{PRICE_POINTS}}
Produce:
1. A comparison table whose rows are the 6–8 attributes that actually drive the choice (not every spec) and whose columns are the models, with values exactly as in my specs.
2. A "Choose this if…" line under each model, written as a use case ("Choose the 1.5 L if you brew for one or two people").
3. What the extra money buys at each step up, one line per step. If a step up adds little, say so.
4. A short paragraph for the product pages that helps a shopper torn between two neighboring models.
5. Label text for the middle option if the data supports it. "Most versatile" is fine; "Most popular" or "Best seller" only if I've given sales data that shows it.
Rules: no "best" ranking without a stated criterion. Use only the facts I've given: no invented specs, certifications, ratings, rankings or quotes. List anything missing under "Facts to confirm".
Fill in: RANGE_DETAILS BUYER_TYPES PRICE_POINTS
List every model's full specs and the real price. If a step up adds little, the prompt will say so, which builds trust.
Best for: Bundles and kits where the saving must be real and checkable
You are an e-commerce merchandiser. Write the product page copy for a bundle or kit.
What's in the bundle (each item, quantity, size): {{BUNDLE_CONTENTS}}
The price of each item bought separately on my store: {{INDIVIDUAL_PRICES}}
Bundle price: {{BUNDLE_PRICE}}
Who the bundle is for and the job it does: {{WHO_ITS_FOR}}
Write:
1. A bundle name that says what it does ("Weekend Bike Repair Kit"), not "Bundle 3".
2. A 60–100 word description built around the job, explaining why these items belong together.
3. "What's inside": a checklist with quantities and a one-line benefit for each item.
4. The saving, calculated: the total of the separate prices minus the bundle price, in currency and as a share of the separate total, with the arithmetic shown.
5. A gift-angle line, if it suits.
Rules: state a saving only if my prices support it, and base "Save X" on the prices the items actually sell for separately on my store, never an inflated "worth" or "value" figure. Reference-price claims are regulated in many markets. If I haven't given a separate price for every item, don't state a saving; tell me what's missing. Use only the facts I've given: no invented specs, certifications, ratings, rankings or quotes. List anything missing under "Facts to confirm".
Fill in: BUNDLE_CONTENTS INDIVIDUAL_PRICES BUNDLE_PRICE WHO_ITS_FOR
Use the prices the items actually sell for separately on your store. The saving is calculated, not estimated.
Review Mining, FAQs & Objections
Your reviews already contain the words customers search with and the doubts that stop the next sale. These prompts mine pasted reviews for phrases and outcomes, sort complaints into fixable expectation gaps and real defects, and handle the price objection with facts you can prove. They also compare your listing with competitors' listings and low-star reviews to find gaps you can honestly win on, and write shipping and returns microcopy. Nothing here writes reviews or testimonials for you.
Best for: Finding the words customers use so your copy sounds like them
You are a conversion copy researcher. Mine my real customer reviews for the words, outcomes and doubts I should use in my product copy.
Product: {{PRODUCT_NAME}}
Reviews (full text, with star ratings; reviewer names removed): {{PASTED_REVIEWS}}
My current product page copy: {{CURRENT_COPY}}
Produce:
1. The outcomes customers describe, ranked by how many reviews mention each, with the count and two short verbatim snippets (quote exactly; don't tidy the grammar).
2. Phrases customers use for the product and its benefits that my copy doesn't use. These are often the words they search with.
3. Moments of use: when, where and why people use it, from the reviews.
4. Doubts and complaints, ranked by count and split into product problems (fix the product), expectation gaps (fix the copy or photos) and delivery or service issues.
5. Gaps between my copy and the reviews: what customers value that the page doesn't mention, and claims in my copy that no reviewer confirms.
6. Five copy suggestions that use customers' own language.
Rules: count, don't estimate, and work only from the reviews I pasted. To quote a review on the page, use it exactly, attributed, and only where the platform allows. Never write new reviews or testimonials, and never merge reviews into an invented quote. Fake reviews are unlawful in the US and UK.
Fill in: PRODUCT_NAME PASTED_REVIEWS CURRENT_COPY
Paste full review text with star ratings and remove reviewer names. Works best with 10 or more reviews.
Example output · Claude (example written by The Prompt Index)
Sample inputs: PRODUCT_NAME = 1 L stovetop gooseneck pour-over kettle. PASTED_REVIEWS = 12 reviews:
R1 5★ "The spout is the whole point. I can finally pour slowly in circles and my pour-overs taste less bitter. Looks great on the stove too."
R2 5★ "Bought this to replace a regular kettle for my V60. Pour control is excellent, slow steady stream. Handle stays cool enough to hold."
R3 4★ "Lovely kettle, precise pour. Only downside is there's no thermometer so I'm guessing water temp."
R4 2★ "Handle gets hot on a gas burner if the flame goes up the side. Burned my fingers twice. Pour is nice though."
R5 5★ "Works on my induction hob, which my last gooseneck didn't. Boils quickly, pours like a dream."
R6 3★ "Smaller than I expected. 1 litre but you can only fill to about 800ml before it spits from the spout."
R7 5★ "Morning ritual upgrade. Slow, controlled pour, and it's pretty enough to leave out on the counter."
R8 4★ "Great pour control. Wish it had a temperature gauge."
R9 1★ "Lid falls off when you tip it to pour the last bit. Annoying."
R10 5★ "Precise pour, heats fast on induction, feels solid. My partner says the coffee is better, which I'll take."
R11 4★ "Does what it says. Gooseneck makes pour-over easy for a beginner like me."
R12 2★ "Says 1L but max fill line is lower. Felt misled."
CURRENT_COPY = "Our 1L gooseneck kettle brings barista precision to your kitchen. Premium stainless steel. Elegant design. Compatible with all hob types."
1. Outcomes customers describe (by number of reviews)
- The pour (control, precision or ease): 9 of 12 (R1, R2, R3, R4, R5, R7, R8, R10, R11). "pour slowly in circles" (R1); "slow steady stream" (R2)
- Works on induction: 2 (R5, R10). "Works on my induction hob, which my last gooseneck didn't" (R5)
- Heats quickly: 2 (R5, R10). "Boils quickly" (R5); "heats fast on induction" (R10)
- Better-tasting coffee: 2 (R1, R10). "my pour-overs taste less bitter" (R1); "My partner says the coffee is better" (R10)
- Good-looking enough to display: 2 (R1, R7). "Looks great on the stove too" (R1); "pretty enough to leave out on the counter" (R7)
- Easy for beginners: 1 (R11). "makes pour-over easy for a beginner like me"
2. Customer phrases your copy doesn't use
"pour control", "slow, controlled pour", "precise pour", "pours like a dream", "works on my induction hob", "morning ritual", "beginner", "V60". Customers talk about control and slowness; your copy says "barista precision", which nobody used.
3. Moments of use
Morning coffee as a ritual (R7); brewing with a V60 or other pour-over dripper (R1, R2); learning pour-over for the first time (R11).
4. Doubts and complaints (by count)
- Expectation gaps (fix the copy or photos): usable capacity below 1 L, 2 (R6, R12); no built-in thermometer, 2 (R3, R8).
- Product problems (raise with your supplier): handle heats up on gas when the flame climbs the side, 1 (R4); lid falls off when tipped to pour the last of the water, 1 (R9).
- Delivery or service issues: none mentioned.
5. Gaps between your copy and the reviews
- Not on the page but valued: induction compatibility, pour control and speed of heating.
- Claims no reviewer confirms: "Compatible with all hob types". Reviewers confirm induction (R5, R10) and gas (R4), but one gas user burned their fingers. Check this claim, and add a line about keeping the flame under the base.
- "Premium stainless steel" isn't mentioned by any reviewer. That's not proof it's wrong, but it is evidence that it isn't what persuades people.
6. Copy suggestions in customers' language
1) Headline: "A slow, steady pour you control, for better pour-over at home."
2) Induction line: "Works on induction, gas and electric hobs." (Add only the hob types you have confirmed; reviews confirm induction and gas.)
3) Capacity, stated honestly: "1 L total capacity; fill to the max line (about 800 ml) for a clean pour." Measure the max line yourself before publishing.
4) Set expectations: "No built-in thermometer. Pair it with a clip-on thermometer if you brew to an exact temperature."
5) Beginner reassurance: "If you're new to pour-over, the narrow spout does the hard part: a thin, steady stream."
Use review quotes on the page only exactly as written and attributed, where your review platform allows it.
Best for: Separating fixable expectation gaps from real product defects
You are an e-commerce product and listing analyst. Find out which complaints my listing could prevent by setting better expectations, and which are real product problems.
1–3 star reviews (full text): {{NEGATIVE_REVIEWS}}
Return reasons and notes from my returns data: {{RETURN_REASONS}}
Current listing (title, bullets or description, image list): {{CURRENT_LISTING}}
Do this:
1. Cluster the complaints by theme, with counts and one exact example quote each.
2. Classify each cluster: expectation gap (the listing implied something the product doesn't do), product defect, sizing or fit, damaged in transit, or misuse.
3. For every expectation gap, write the specific listing fix: a sentence to add, a claim to soften, a photo to add (such as a scale shot with a hand) or a "Not for…" line.
4. For misuse, write the instruction or warning that would have prevented it.
5. List the product defects separately, with counts, for my supplier conversation. Don't hide them in copy.
6. Rank all fixes by how many complaints each could prevent.
Rules: don't write copy that argues with reviewers or hides a real defect. Use only the reviews and returns I pasted, and count rather than estimate.
Fill in: NEGATIVE_REVIEWS RETURN_REASONS CURRENT_LISTING
Paste your 1–3 star reviews and return reasons together. Defects go to a supplier list instead of being hidden in copy.
Best for: Premium DTC products that lose sales on price
You are a direct-response copywriter for a DTC brand. My product costs more than the alternatives shoppers compare it with. Write page copy that handles the price objection honestly.
Product: {{PRODUCT_NAME}}
Our price: {{OUR_PRICE}}
What shoppers compare us with (product types or price points; no competitor names needed): {{ALTERNATIVES}}
Verifiable differences (materials, warranty length, testing, where it's made, cost per use): {{REAL_DIFFERENCES}}
Objections we hear (support, reviews, social comments): {{OBJECTIONS}}
Write:
1. A "Why it costs more" block (80–120 words) that explains where the money goes, using only the verifiable differences.
2. A cost-per-use or cost-over-time comparison, only if my facts support it, with the arithmetic shown and the assumptions labeled (for example "if used daily for 3 years").
3. An answer to each objection in 2–3 sentences, in the customer's language.
4. A risk-reversal line using my real returns or warranty terms; if I have none, skip it.
5. Three short trust lines for near the add-to-cart button, each based on a fact I gave.
Rules: no disparaging claims about named competitors, and no comparative claims ("lasts twice as long") unless I supplied the test behind them. Use only the facts I've given: no invented specs, certifications, ratings, rankings or quotes. List anything missing under "Facts to confirm".
Fill in: PRODUCT_NAME OUR_PRICE ALTERNATIVES REAL_DIFFERENCES OBJECTIONS
List only differences you can prove: materials, warranty length, testing, origin. Comparative claims need your own test data.
Best for: Delivery and returns lines near the price, in the cart and in an accordion
You are a UX writer for e-commerce. Turn my policies into short, honest microcopy for the product page and cart, where shoppers decide.
Shipping policy (carriers, costs, free-shipping threshold, dispatch cut-off, delivery estimates): {{SHIPPING_POLICY}}
Returns policy (window, condition, who pays, exclusions): {{RETURNS_POLICY}}
Warranty terms: {{WARRANTY_TERMS}}
Regions we ship to: {{STORE_REGIONS}}
Write:
1. A delivery line under the price (max 12 words): the order-by cut-off and estimated arrival as a range, never a promised date unless my policy commits to one.
2. Text for three trust badges (max 4 words each), based only on my policies.
3. A "Delivery & Returns" accordion tab of 60–100 words in plain English, with exclusions included rather than hidden.
4. A cart-page reassurance line (max 15 words).
5. A free-shipping nudge for the cart ("You're [AMOUNT] away from free delivery"), only if I have a threshold.
6. A list of mismatches or ambiguities in my policies that a shopper could read two ways.
Rules: every line must match the policy exactly. Say "free returns" only if returns are free in every case I listed; otherwise say what is free. Statutory rights, such as the 14-day cancellation right for most online purchases in the UK and EU, can't be reduced by your copy; flag any policy wording that seems to. This is not legal advice.
Fill in: SHIPPING_POLICY RETURNS_POLICY WARRANTY_TERMS STORE_REGIONS
Paste your actual policies. The prompt also lists policy wording a shopper could read two ways.
Best for: Finding honest angles where competitor listings and products fall short
You are a marketplace research analyst. Compare my listing with the competitors shoppers see next to it and find the gaps I can honestly win on.
My listing (title, bullets or description, price, image list): {{MY_LISTING}}
2–4 competitor listings, pasted as text (title, bullets, price, image list): {{COMPETITOR_LISTINGS}}
Their 1–3 star reviews, pasted with reviewer names removed: {{COMPETITOR_REVIEWS}}
Produce:
1. A feature and claim matrix: rows are the attributes buyers compare, columns are my listing and each competitor, and each cell holds what that listing actually states ("not stated" where it's silent).
2. The complaint themes in the competitors' reviews, with counts and one short quote each.
3. For each theme, whether my product genuinely avoids that problem: "yes, per my facts", "unknown, test it" or "no". Only "yes" themes become copy.
4. Questions none of the listings answer. These are easy wins for my bullets, images or Q&A.
5. Five specific edits to my listing, each tied to a gap above.
Rules: don't name or disparage competitors in my copy, and don't claim superiority ("unlike other brands, ours never leaks") unless I've tested it. Use only the text I pasted, and count rather than estimate. If something about my own product is unclear, ask me instead of assuming.
Fill in: MY_LISTING COMPETITOR_LISTINGS COMPETITOR_REVIEWS
Paste 2–4 competitor listings and their low-star reviews as text. Only gaps your product genuinely closes become copy.
Email, Abandoned Cart & Retention
Email is where a store earns its second order. These prompts write abandoned cart and browse abandonment flows, a post-purchase series that heads off returns and asks for reviews the compliant way, back-in-stock alerts, launch sequences and replenishment emails timed to real usage. None of them use countdowns that reset or 'only 2 left' unless the number is live.
Best for: Recovering carts from shoppers who gave an email but didn't check out
You are an e-commerce lifecycle email copywriter. Write a three-email abandoned cart sequence for shoppers who left items in their cart after giving their email.
Store and what we sell: {{STORE_DESCRIPTION}}
Brand voice: {{BRAND_VOICE}}
Shipping, returns and payment options (true facts only): {{POLICIES}}
Incentive rules (for example none, or free shipping in email 3 only): {{INCENTIVE_RULES}}
Main reasons people hesitate in our category: {{TOP_OBJECTIONS}}
For each email give the send timing, three subject lines (max 45 characters), preview text (max 90 characters), a 50–120 word body with a merge tag for the cart contents such as [CART_ITEMS], and one button label.
- Email 1 (about 1 hour): a helpful reminder, no discount.
- Email 2 (about 24 hours): answer the top objection with our real policies. Social proof only through a merge tag for genuine reviews ([PRODUCT_REVIEW]), never written by you.
- Email 3 (about 48–72 hours): a last reminder, with the incentive only if my rules allow it.
Rules: no fake urgency ("your cart expires", "only 2 left") unless the system really does that; show stock levels only through a live inventory merge tag. Include an unsubscribe line and a placeholder for my business address, as anti-spam laws (CAN-SPAM, UK and EU rules) require. Add a 160-character SMS version of email 1 for SMS opt-ins.
Fill in: STORE_DESCRIPTION BRAND_VOICE POLICIES INCENTIVE_RULES TOP_OBJECTIONS
Give your real policies and incentive rules. Merge tags in square brackets are placeholders for your email platform's own tags.
Best for: Subscribers who viewed products but never added anything to cart
You are a lifecycle marketer. Write a browse abandonment email for subscribers who viewed a product page in the last day but didn't add anything to their cart.
Category they browsed: {{PRODUCT_CATEGORY}}
Brand voice: {{BRAND_VOICE}}
Helpful resources we have (size guide, buying guide, quiz, comparison chart, live chat): {{HELP_RESOURCES}}
What shoppers in this category usually need to decide (fit, compatibility, material, gift suitability): {{DECISION_FACTORS}}
Write two versions:
A) "Still deciding?": helps them choose, using one resource and a short tip.
B) "You might also like": shows the viewed product plus alternatives through merge tags ([VIEWED_PRODUCT], [RECOMMENDED_PRODUCTS]).
For each version: three subject lines (max 45 characters), preview text, a 60–100 word body, a button label, and a P.S. that answers one decision factor.
Rules: browse emails can feel intrusive, so never write "we saw you looking at…" or anything that sounds like surveillance; use something like "Still thinking it over?". Send only to people who agreed to marketing email, and include the unsubscribe line. No discount by default: browsing isn't commitment, and discounting here teaches people to browse and wait. Tell me which version to test first, and what result would make me switch.
Fill in: PRODUCT_CATEGORY BRAND_VOICE HELP_RESOURCES DECISION_FACTORS
List the help resources you really have (size guide, quiz, chat). Send only to people who agreed to marketing email.
Best for: Helping new customers succeed with the product and earning honest reviews
You are a retention email strategist. Write a post-purchase email series for {{PRODUCT_NAME}} that helps customers succeed with the product, cuts avoidable returns and earns honest reviews.
How to set up or start using it: {{SETUP_STEPS}}
Common problems and their fixes (from support tickets): {{COMMON_ISSUES}}
Where we ask for reviews (our site, Amazon, Trustpilot, Etsy): {{REVIEW_PLATFORM}}
A genuinely related product to suggest later: {{CROSS_SELL}}
Write four emails, each with timing, subject line, preview text and 60–120 words:
1. Order confirmed: what happens next (dispatch, tracking).
2. Delivered: get the most from it in three steps, plus the fix for the most common problem.
3. Seven to fourteen days after delivery, depending on how soon the product shows its value: a check-in that offers help, then a neutral review request.
4. About 30 days: a care tip and the related product, framed around use.
Review rules: ask every customer the same way, whatever their likely opinion. Don't select happy customers, offer incentives, or steer toward five stars. On Amazon, use only neutral requests (its Request a Review button or neutral wording), with no incentives and no "contact us instead of reviewing". Offer support alongside the review link, never instead of it.
Mark which emails are transactional and which are marketing (needing unsubscribe text), for me to check against local law.
Fill in: PRODUCT_NAME SETUP_STEPS COMMON_ISSUES REVIEW_PLATFORM CROSS_SELL
Pull common problems from support tickets. The review request follows marketplace and consumer-protection rules.
Best for: Restock alerts and low-stock messages based on real inventory
You are an e-commerce email copywriter. Write back-in-stock and low-stock messages for {{PRODUCT_NAME}} that create real urgency from real stock, never invented scarcity.
Restock facts (date, quantity, which variants, any changes): {{RESTOCK_DETAILS}}
Brand voice: {{BRAND_VOICE}}
What stock data my platform can insert live (for example a quantity merge tag, variant availability): {{STOCK_DATA}}
Write:
1. A back-in-stock email to people who signed up for alerts: three subject lines (max 40 characters), preview text, a 50–80 word body and a button label. Say which variants are back and, if my facts include it, how long the last restock lasted, worded as a fact, not a threat.
2. A 160-character SMS version for opted-in subscribers.
3. A low-stock line for the product page and one for email that use the live quantity merge tag. If my platform can't insert live stock, write versions without numbers.
4. A pre-order or "join the waitlist" version for when it sells out again, with the expected date stated as an estimate.
Rules: no "only 3 left" unless that is the live number, no countdown timers that reset, and no "selling fast" without data. Consumer regulators such as the FTC in the US and the CMA in the UK treat fake scarcity as a misleading practice, and it erodes trust. Include unsubscribe text in marketing emails.
Fill in: PRODUCT_NAME RESTOCK_DETAILS BRAND_VOICE STOCK_DATA
Check what live stock data your platform can insert. Without it, the prompt writes versions with no numbers.
Best for: Launching a new product to existing subscribers and past buyers
You are a DTC launch copywriter. Write a launch email sequence for {{PRODUCT_NAME}} to my existing subscribers.
Launch facts (what it is, why we made it, price, quantities, what's new compared with our current range): {{LAUNCH_FACTS}}
Who's on the list and what they've bought before: {{AUDIENCE}}
Launch offer, if any, with exact terms and end date: {{LAUNCH_OFFER}}
Launch date and time zone: {{LAUNCH_DATE}}
Write four emails, each with send timing, two subject lines (max 45 characters), preview text, an 80–150 word body and a button label:
1. Teaser (3–5 days before): the problem it solves and a behind-the-scenes detail, with an early-access sign-up if we offer one.
2. Launch day: what it is, who it's for, the key facts and the offer terms.
3. Two or three days later: answers to the questions people will have (sizing, compatibility, how it differs from our current product).
4. Offer ending (only if there's a real end date): a short reminder with the exact end time.
Add a plain-text version of email 2.
Rules: if quantities are limited, state the real number; if they aren't, don't imply it. Quote customer reactions only once real reviews exist. Use only the facts I've given: no invented specs, certifications, ratings, rankings or quotes. List anything missing under "Facts to confirm".
Fill in: PRODUCT_NAME LAUNCH_FACTS AUDIENCE LAUNCH_OFFER LAUNCH_DATE
State the launch offer's exact terms and end date. If stock isn't limited, the copy won't imply it is.
Best for: Consumables, refills and accessories with a predictable reorder cycle
You are a lifecycle strategist for a consumables or accessories brand. Plan and write replenishment and cross-sell emails timed to when customers actually run out or need the next item.
Product: {{PRODUCT_NAME}}
How long one unit lasts at typical use (from product facts or my reorder data): {{USAGE_RATE}}
Products that genuinely go with it: {{RELATED_PRODUCTS}}
Reorder data I have (median days between orders, if known): {{ORDER_DATA}}
Do this:
1. Work out the send timing: a reminder a few days before the expected run-out, based on my data or the usage rate. Show the arithmetic and the source.
2. Write the replenishment email: subject line options, preview text, a 50–90 word body, a one-click reorder button label, and a subscribe-and-save mention only if I offer it.
3. Write a cross-sell email for each related product: why it pairs with what they bought, in terms of use, in 60–100 words.
4. Give suppression rules for these emails: already reordered, returned the product, or has an open support ticket.
5. The one variable to test first.
Rules: don't state that the product "runs out in X days" as a fact for everyone; say "about" and base it on my data. Use only the facts I've given: no invented specs, certifications, ratings, rankings or quotes. List anything missing under "Facts to confirm".
Fill in: PRODUCT_NAME USAGE_RATE RELATED_PRODUCTS ORDER_DATA
Use your median days between orders if you have it; otherwise the usage rate from the product facts. The timing math is shown.
Ads, Product Photos & AI Images
Ad copy for Meta and Google Search, an Amazon PPC search term triage, short-form video scripts with proper creator disclosure, and a photo brief that meets Amazon's main image rules. Two prompts work with AI image editors: one builds lifestyle scenes around your real product photo, and one cleans up phone photos without changing what the customer will receive. An infographic plan turns your secondary images into answers.
Best for: Facebook and Instagram ads where you test angles, not word swaps
You are a performance creative strategist for DTC brands. Write Facebook and Instagram ad copy for {{PRODUCT_NAME}}, built so I can test angles, not just word swaps.
Product facts: {{PRODUCT_FACTS}}
Who we're targeting and what they care about: {{AUDIENCE}}
Current offer, with exact terms (or "none"): {{OFFER}}
Customer phrases from real reviews (optional): {{REVIEW_PHRASES}}
Write five angles: problem and solution, outcome, objection-buster, gift, and "the old way vs our way" (no competitor names). For each:
1. Primary text: a short version (125 characters or fewer, shown before "See more") and a longer version (60–100 words).
2. A headline (40 characters or fewer) and a description (30 characters or fewer).
3. The image or video idea that matches the angle, in one line.
4. The hypothesis: who this angle should win with, and why.
Rules: quote customer phrases only exactly as they appear in real reviews; don't create testimonials. Don't assert or imply personal attributes in the way Meta's ad policies prohibit ("Are you overweight?", "Struggling with debt?"). No before-and-after or health claims for regulated products, no "guaranteed results" and no invented discounts. Tell me which angle you'd launch first on a small budget.
Fill in: PRODUCT_NAME PRODUCT_FACTS AUDIENCE OFFER REVIEW_PHRASES
Bring real review phrases if you have them. Launch the angle the prompt recommends on a small budget first.
Best for: Search campaigns pointing at a product or collection page
You are a Google Ads specialist for e-commerce. Write a responsive search ad for {{PRODUCT_OR_CATEGORY}}.
Landing page facts (what the page sells, prices, delivery, returns, range size): {{LANDING_PAGE_FACTS}}
Keywords in this ad group: {{KEYWORDS}}
Current offer, with exact terms (or "none"): {{OFFER}}
Produce:
1. 15 headlines of 30 characters or fewer, with character counts: 3 with the main keyword, 3 on product benefits, 3 on reassurance (delivery, returns), 2 on the offer (only if it's real), 2 calls to action and 2 brand headlines.
2. Four descriptions of 90 characters or fewer.
3. Which 2–3 headlines to pin, if any, and why. Pin sparingly, because pinning reduces the combinations Google can test.
4. Four sitelink ideas with link text of 25 characters or fewer and two description lines of 35 characters or fewer each, pointing to page types I probably have (size guide, returns, new arrivals). Mark them as suggestions to match to my real URLs.
5. Six callouts of 25 characters or fewer, using true facts only.
Rules: every claim must appear on the landing page. No exclamation marks in headlines, no ALL CAPS words, no repeated punctuation, and no superlatives ("best", "#1") unless the page shows third-party proof.
Fill in: PRODUCT_OR_CATEGORY LANDING_PAGE_FACTS KEYWORDS OFFER
Paste what the landing page actually says. Ads that promise what the page doesn't say get disapproved and waste clicks.
Best for: TikTok, Reels and Shorts product videos filmed by you or a creator
You are a short-form video strategist for e-commerce. Write scripts for a 20–35 second vertical video about {{PRODUCT_NAME}} for {{PLATFORM}}, to be filmed by {{CREATOR_TYPE}} (me, a staff member or a paid creator).
Product facts: {{PRODUCT_FACTS}}
The objection that stops people buying: {{KEY_OBJECTION}}
Write three scripts with different hooks: a demonstration hook (show it working in the first 2 seconds), a problem hook and an objection hook. For each:
1. Hook: the first line spoken, plus what's on screen in the first 2 seconds.
2. Shot list: 4–7 shots, each with its duration, what to film and the on-screen text (max 6 words).
3. Voiceover or talking-to-camera lines that sound natural, not read.
4. An ending with a call to action that fits the platform.
5. A caption (150 characters or fewer) and 3–5 relevant hashtags.
Rules: the demo must show the real product doing what the facts say, with no effects or edits that exaggerate results. A creator who is paid or received the product free must disclose it (the platform's paid-partnership label plus a clear tag such as #ad), as FTC and UK ASA rules require. Don't script results the creator hasn't had; they speak from their own use. No health, earnings or before-and-after claims for regulated products.
Fill in: PRODUCT_NAME PLATFORM CREATOR_TYPE PRODUCT_FACTS KEY_OBJECTION
Tell it who will film. Paid or gifted creators must disclose the relationship, and the prompt builds that in.
Best for: Planning a product shoot that answers buyer questions and meets channel rules
You are an e-commerce art director. Write a shot list and photography brief for {{PRODUCT_NAME}} so the images answer buyers' questions and meet each channel's rules.
Channels the photos are for (Shopify, Amazon, Etsy, eBay, ads): {{CHANNELS}}
Product details (size, materials, colors or variants, key features, packaging): {{PRODUCT_DETAILS}}
Questions buyers ask that photos could answer (scale, texture, what's included, how it's used): {{BUYER_QUESTIONS}}
Budget and setup (phone and lightbox, studio, location): {{BUDGET_AND_SETUP}}
Deliver:
1. A numbered shot list of 7–10 images in gallery order: main image, angles, detail or texture, scale (with a hand or a common object), in use, what's in the box, variants, and a clean shot for infographic text. For each: purpose, framing, background, props and a lighting note.
2. Channel requirements: Amazon's main image needs a pure white background (RGB 255, 255, 255), the product filling about 85% of the frame, and no text, logos, watermarks or extra props; I'll confirm the current style guide for my category. Note anything Etsy or eBay require.
3. Technical specs: resolution (at least 1,000 px on the longest side for Amazon zoom), aspect ratio per channel, and file naming.
4. A styling note so colors are true to life (white balance, no heavy filters).
5. A "don't shoot" list: anything that would misrepresent size, contents or condition.
Fill in: PRODUCT_NAME CHANNELS PRODUCT_DETAILS BUYER_QUESTIONS BUDGET_AND_SETUP
List the questions buyers ask that a photo could answer. Confirm the image rules in your category's current style guide.
Best for: Lifestyle and seasonal scenes without a location shoot
You are an AI product-photography director. Write image-generation prompts for lifestyle and background scenes for my product. I'll use a real photo of the product as the input image, so the product stays accurate.
The product exactly as it looks (shape, color, material, size next to a reference object, label details): {{PRODUCT_DESCRIPTION}}
Scenes I want (room, season, moment of use, props): {{SCENES_WANTED}}
Brand look (colors, mood, lighting): {{BRAND_LOOK}}
The image tool I'm using (an editor that keeps my product photo, or a general image model): {{IMAGE_TOOL}}
Where the images will run (product page, ads, social) and the aspect ratio: {{CHANNEL}}
Write six prompts. Each includes the scene, camera angle and lens feel, lighting, surface and props, where the product sits in the frame, negative space for text if it's an ad, and an instruction to keep the product's shape, color, label and proportions exactly as in the reference photo.
Then add a checklist for comparing each output with the real product (scale, color, logo text, number of parts), and say which scenes suit gallery images and which suit ads.
Rules: never use an AI image as the main product image, or to show features, colors or sizes the product doesn't have. Don't show the product used in ways it can't be used. Label AI imagery wherever a platform requires it.
Fill in: PRODUCT_DESCRIPTION SCENES_WANTED BRAND_LOOK IMAGE_TOOL CHANNEL
Use these with an image editor that takes your real product photo as the input. Compare every result with the real product before publishing.
Example output · Claude (example written by The Prompt Index)
Sample inputs: PRODUCT_DESCRIPTION = handmade stoneware mug, 12 oz, 9.5 cm tall, 8.5 cm wide, matte speckled oatmeal glaze with a glossy unglazed-look rim band, hand-pulled C-shaped handle, small stamped logo on the base only (not visible from the side). SCENES_WANTED = morning kitchen window, desk on an autumn afternoon, weekend breakfast flat lay, winter evening by a window, gift being unwrapped, outdoor café table. BRAND_LOOK = warm neutrals, soft natural light, linen and pale oak, calm and uncluttered. IMAGE_TOOL = an image editor that keeps my uploaded product photo and generates the scene around it. CHANNEL = Shopify gallery (1:1) and Instagram feed ads (4:5).
Prompt 1: Morning kitchen window (gallery, 1:1)
"Place the mug from the reference photo on a pale oak windowsill in a calm kitchen, early morning. Soft side light from the window on the left, gentle shadow falling right. Eye-level camera, 50 mm lens feel, shallow depth of field so the background is a soft blur of white tiles and a single green plant. Steam rising lightly from black coffee in the mug. Keep the mug's shape, size, speckled oatmeal glaze, rim band and handle exactly as in the reference photo. Don't add a logo or text to the mug."
Prompt 2: Autumn desk (ad, 4:5)
"Place the reference mug on a light oak desk next to a closed notebook and a pencil, late afternoon in autumn. Low golden light from the right, long soft shadows. Three-quarter view from slightly above, 35 mm lens feel. Leave the top third of the frame as plain warm-beige wall for ad text. Keep the mug identical to the reference: same proportions, matte speckled glaze, handle shape and rim band. Mug at actual scale next to the notebook."
Prompt 3: Weekend breakfast flat lay (gallery, 1:1)
"Overhead flat lay on natural linen: the reference mug filled with milky coffee, a plate with a croissant, a small jug of milk and a folded napkin. Diffused daylight with no harsh shadows. The mug sits at the lower-right third with its handle pointing right. Keep the mug's glaze color, speckle pattern and proportions exactly as in the reference. True-to-life scale, so the mug is visibly smaller than the plate."
Prompt 4: Winter evening by a window (ad, 4:5)
"The reference mug held loosely in two hands wearing a cream knitted sweater, seen from the chest down. Blue-hour light through a window behind, a warm lamp glow from the left. 85 mm lens feel, background softly blurred. Keep the mug's size realistic in adult hands (9.5 cm tall), and its glaze, rim and handle identical to the reference. Leave space at the bottom of the frame for a caption."
Prompt 5: Gift being unwrapped (gallery, 1:1)
"The reference mug partly lifted out of plain kraft-paper tissue in an open, undyed cardboard box on a pale oak table. Soft overhead daylight. Hands just entering the frame from the bottom. Keep the mug exactly as in the reference photo. Don't add branded packaging, ribbons with text or a logo to the box."
Prompt 6: Outdoor café table (ad, 4:5)
"The reference mug on a small round marble café table outdoors, morning, with a softly blurred street and greenery behind. Natural overcast light. Eye-level, 50 mm lens feel. Leave negative space on the left for ad text. Keep the mug's shape, glaze and handle exactly as in the reference, at true scale next to a teaspoon on a saucer."
Before you publish, check each result against the real mug:
- Height and width look right against the props (a teaspoon is roughly 13–15 cm long).
- Glaze color matches the real mug in daylight, not the warmer scene light. Compare side by side.
- The speckle pattern and rim band are still there and haven't been smoothed away.
- The handle shape is unchanged. Image models often redraw handles.
- No logo, text or decoration has appeared on the side of the mug.
- Any coffee or steam isn't hiding a defect in the glaze.
Gallery vs ads: prompts 1, 3 and 5 suit gallery images, because they show scale, use and the gift moment. Prompts 2, 4 and 6 suit ads, because they leave room for text. Your main product image should stay a real photo of the mug on a plain background. Prompt 4 shows a person's hands, so check that they hold the mug in a natural way and in proportion.
Best for: Secondary images that answer questions on mobile before shoppers read bullets
You are an Amazon listing designer. Plan the text and layout for my secondary images (infographics), which many shoppers on mobile see before they read a single bullet.
Product facts and specs: {{PRODUCT_FACTS}}
Top buyer questions, from reviews and Q&A: {{TOP_QUESTIONS}}
What competitor listings in my category fail to show: {{COMPETITOR_GAPS}}
Number of secondary image slots I'll use: {{IMAGE_COUNT}}
For each image slot, give:
1. The job of the image: the one question it answers (size, what's included, how it works, materials, compatibility, care).
2. Headline text (max 6 words) and up to three callouts (max 5 words each), large enough to read on a phone.
3. The visual: photo or render, angle, annotations, and any arrows or dimension lines.
4. The fact from my list that each piece of text relies on.
Also give the order of the images, a dimensions image with my exact measurements in both metric and imperial, and a text-free alternative for the in-use shot.
Rules: no badges or claims I can't support ("#1", "Best Seller", "Amazon's Choice", "Eco", "Clinically proven", or certification logos I don't hold), and nothing that imitates Amazon's own badges. Keep text to one idea per image. Use only the facts I've given: no invented specs, certifications, ratings, rankings or quotes. List anything missing under "Facts to confirm".
Fill in: PRODUCT_FACTS TOP_QUESTIONS COMPETITOR_GAPS IMAGE_COUNT
List top buyer questions and what competitor images fail to show. Your designer gets one idea per image.
Best for: Cleaning up phone photos without changing what the customer will receive
You are a product photo retoucher. I've taken product photos on my phone and want to fix them with an AI image editor without making the product look different from what a customer will receive.
What the photo shows: {{PHOTO_DESCRIPTION}}
Problems I see (shadows, color cast, cluttered background, dust, crooked angle, reflections): {{PROBLEMS}}
Where it will be used (Amazon main image, Shopify gallery, Etsy thumbnail, ad): {{TARGET_USE}}
Write:
1. A sequence of edit instructions with one change per step, because editors handle one clear change better than five at once. Phrase each as a prompt I can paste, for example: "Replace the background with pure white (#FFFFFF). Keep the product and its natural contact shadow unchanged."
2. For each step, what to check in the result before moving on.
3. Two clearly separated lists: acceptable edits (background, dust, lighting balance, straightening, removing props) and edits that misrepresent the product (changing its color, removing scratches from a used item, smoothing a texture, adding features).
4. Export settings for the target use: size, aspect ratio, file type and sRGB color profile.
Rules: color must match the real item in daylight. If a correction might change how the color looks, tell me to compare it side by side with the product. For used or refurbished items, flaws stay visible.
Fill in: PHOTO_DESCRIPTION PROBLEMS TARGET_USE
Make one edit at a time and check each result. Used and refurbished items keep their flaws visible.
Best for: Cutting wasted Sponsored Products spend and harvesting converting terms
You are an Amazon Ads analyst. Triage my Sponsored Products search term report so I stop paying for irrelevant clicks and move proven terms into their own targeting.
Search term report export, at least 30 days, with all columns: {{SEARCH_TERM_REPORT}}
My target or break-even ACoS, and how I worked it out: {{TARGET_ACOS}}
What the product is and isn't: {{PRODUCT_FACTS}}
How my campaigns are structured now: {{CAMPAIGN_STRUCTURE}}
Do this:
1. Calculate ACoS, CPC and conversion rate for each term, showing each formula once. If you can run code, use it.
2. Sort the terms into groups: harvest (converting within target, so move to exact match), negate exact (clicks above a threshold with no orders; state the threshold you used and why), negate phrase (wrong meaning or wrong product type), watch (too little data), and fix the listing (relevant, plenty of clicks, poor conversion, so the page may be the problem rather than the ad).
3. A table per group: term | clicks | spend | orders | ACoS | action.
4. A bid direction and range for harvested terms, with the reasoning.
5. Search terms whose wording my title or bullets should use.
Rules: use only my numbers, and don't quote what a "good" ACoS is for my category. Don't negate my brand terms or anything that has converted without telling me why.
Fill in: SEARCH_TERM_REPORT TARGET_ACOS PRODUCT_FACTS CAMPAIGN_STRUCTURE
Export at least 30 days of the search term report. Low-data terms are flagged to watch, not negated early.
How to get the best results
1. Paste facts, not adjectives
Every prompt has {{VARIABLES}} in capitals. Replace each one, braces included, with real product data: the spec sheet, the label text, the measurement chart, the actual returns policy. "Stainless steel, 710 ml, hot for up to 8 hours in the manufacturer's test" gets you a usable page. "High-quality bottle" gets you filler. If you don't know a fact, leave it out; the prompts list missing facts instead of guessing.
2. Set up your voice once
Run the Brand Voice Card prompt first and save the result. Paste it wherever a prompt asks for brand voice, so a catalog written over months still sounds like one store. In ChatGPT or Claude, keep it in a Project with your policies and size charts attached.
3. Work in batches and check the output
For bulk work, clean the data first, write 10–25 rows at a time, then run the fact-check prompt on the drafts before anything is published. AI models do sometimes move a feature from one product to the next, and a wrong dimension costs you a return. If you can run code in your AI tool, use it for character counts and byte limits; otherwise count them in your spreadsheet.
4. Keep claims honest
Health, cosmetic, food, environmental and "#1" claims are where listings get suppressed and stores get complaints. Keep evidence for every claim you publish, confirm current marketplace limits in your seller dashboard, and check regulated claims with an adviser for your market.
5. Polish and picture it
To tighten a draft or adjust its tone, paste it into our Writing Assistant. For product shots, use the photo brief and AI image prompts with our Image Tools or any image editor that keeps your real product photo as the input, and compare every result with the actual product before you publish.