12 prompts
AI prompts for ecommerce
Updated
These AI prompts for ecommerce help store owners write product pages, category copy, emails, ads and review replies that sell without sounding generic. Shoppers buy on specifics: materials, sizing, delivery times and real reviews. Paste those details and your actual policies into the [BRACKETS] so nothing in the output is invented.
Product and category pages
Product description from specs
Use it for every new product listing.
You are an ecommerce copywriter for [STORE], which sells to [CUSTOMER]. Voice: [VOICE_NOTES]. Product specs: [NAME, MATERIALS, DIMENSIONS, SIZES, COLORS, CARE, PRICE]. Main use cases: [USE_CASES]. Common customer questions: [QUESTIONS]. Write: a title under 70 characters, a 2-sentence opening that sells the main benefit, 5 benefit-led bullets, a sizing or fit note, and a care section. Do not claim certifications or results not listed in the specs.
Category page intro and FAQ
Use it on collection pages that have products but no copy.
Act as an SEO copywriter for an online store. Category: [CATEGORY]. Target search term: [KEYWORD]. Products in it: [PRODUCT_TYPES_AND_PRICE_RANGE]. Write a 100-word intro that helps shoppers choose, then 4 FAQ items about choosing, sizing, materials and shipping using our policies: [POLICIES]. Use the search term once in the intro naturally.
Product comparison table
Use it when shoppers cannot tell similar products apart.
You are a product merchandiser. Products: [PRODUCT_A_SPECS], [PRODUCT_B_SPECS], [PRODUCT_C_SPECS]. Build a comparison table with the 6 attributes shoppers care about most for [CATEGORY]. Below it, write 'Choose X if...' for each product in one sentence.
Size guide copy
Use it to cut returns caused by fit.
Act as a customer experience writer. Product type: [TYPE]. Measurements: [SIZE_CHART]. Fit notes from returns and reviews: [NOTES]. Write a short how-to-measure guide, a fit note per size, and 3 tips for people between sizes. Plain language, under 200 words.
Email and retention
Abandoned cart sequence
Use it to recover carts with useful reminders, not pressure.
You are a lifecycle marketer for [STORE]. Product category: [CATEGORY]. Reasons shoppers hesitate: [SHIPPING_COST, SIZING, PRICE, TRUST]. Our policies: [SHIPPING, RETURNS, GUARANTEE]. Write 3 emails: 1 hour after (reminder with product image note), 24 hours after (answer the top hesitation), 72 hours after (social proof from real reviews: [REVIEWS]). Subject lines under 45 characters. No fake urgency or countdowns unless an offer really ends: [OFFER_END_DATE_IF_ANY].
Post-purchase flow
Use it to turn first-time buyers into repeat customers.
Act as a retention marketer. Product: [PRODUCT]. Time to first use: [DAYS]. Typical reorder interval: [DAYS]. Write 4 emails: order thank you with what to expect, a how-to-get-the-most email, a review request, and a reorder or cross-sell email with [RELATED_PRODUCTS]. Each under 120 words.
Win-back email
Use it for customers who have not ordered in a while.
You are an email marketer for [STORE]. Segment: customers with no order in [MONTHS] months. What is new since then: [NEW_PRODUCTS_OR_IMPROVEMENTS]. Offer (optional): [OFFER]. Write 2 versions of a win-back email: one leading with what is new, one leading with the offer. Under 130 words each, 3 subject lines each.
Ads and promotions
Meta and TikTok ad angles
Use it when launching ads for a product.
Act as a performance creative strategist. Product: [PRODUCT]. Customer: [CUSTOMER]. Top review quotes: [QUOTES]. Price and offer: [OFFER]. Write 5 ad angles (problem, demo, social proof, comparison, gift). For each: a hook for the first 3 seconds of video, primary text under 125 characters and a headline under 40 characters.
Promo calendar for a quarter
Use it to plan sales without training customers to wait for discounts.
You are an ecommerce marketing manager. Store: [STORE]. Key dates: [DATES]. Margin constraints: [MAX_DISCOUNT]. Launches planned: [LAUNCHES]. Plan 12 weeks of promotions mixing: product launches, bundles, gift with purchase, content-led weeks with no discount, and at most [NUMBER] discount events. Return a table with week, theme, offer, channels and goal.
Google Shopping title optimization
Use it to improve product titles in your feed.
Act as a shopping feed specialist. Current titles and product attributes: [PASTE_FEED_ROWS]. Rewrite each title in this order where relevant: brand, product type, key attribute, material, color, size. Keep under 150 characters and put the most searched terms first. Return a table with old title and new title.
Reviews and support
Review reply drafts
Use it to answer reviews quickly and consistently.
You are the customer care lead at [STORE]. Voice: [VOICE_NOTES]. Policies: [RETURNS_AND_SUPPORT_CONTACT]. Reviews: [PASTE_REVIEWS_WITH_RATING]. Draft a reply to each. Thank specific praise, own problems without excuses, and give a clear next step for unhappy customers. Never promise anything outside our policies. Mark replies that need a human decision.
Review theme analysis
Use it monthly to find product and copy fixes.
Act as a product analyst. Reviews from the last [PERIOD]: [PASTE_REVIEWS]. Group praise and complaints into themes with counts and quotes. Suggest product page copy changes that set better expectations, and product issues to raise with the supplier.
Run these automatically
Thig's Writer agent and SEO agent handle category content and on-page fixes, and the Influencer agent finds creators for product seeding. See how it works for ecommerce brands, and add Product schema with the free schema markup generator.
Frequently asked questions
Can AI write product descriptions that rank?
Yes, if you give it real product specs, materials, sizing, use cases and the terms shoppers search for. Avoid publishing near-identical descriptions across similar products.
What is a good prompt for abandoned cart emails?
Give the model your brand voice, the product, likely reasons for hesitation (shipping, sizing, price) and your real policies, then ask for a short sequence that answers those concerns.
Can I use AI to reply to customer reviews?
AI can draft replies quickly. Read each one before posting, especially for negative reviews, and never promise refunds or fixes you cannot deliver.
