Here’s a compact, practical guide and ready-to-use prompt templates to generate high-quality product, category and CMS content for PrestaShop with any LLM (ChatGPT, etc.). Use these templates as-is or copy/paste into your automation to produce CSV/JSON ready for PrestaShop import.
Core principles (what makes a “perfect” prompt)
- Be explicit: include product data, brand voice, target audience, SEO keywords, and any forbidden claims.
- Specify output format: CSV, JSON, HTML-wrapped text, or plain text for copying into PrestaShop fields.
- Constrain length and s: character limits for meta title/description, short vs long description lengths, required HTML tags.
- Provide accurate data and state when the model must not invent specs (dimensions, weight, materials).
- Ask for SEO elements: meta title, meta description, URL slug (link_rewrite), alt text, suggested tags/keywords.
- Ask for variation: short description (for listings), long description (product page), bullets for features, FAQs, and schema (JSON-LD) if desired.
- Instruct the model to return only the data in the required format — no extra commentary.
System-level instruction (use as ChatGPT “system” or prefix)
You can start with a system message:
"You are a professional e-commerce copywriter and SEO specialist experienced with PrestaShop. Produce concise, accurate, SEO-optimized product/category/CMS content with specified output format. Do not invent technical specifications. Return only the requested structured output."
Product prompt template (single product)
Produce CSV/JSON or PrestaShop columns. Replace placeholders in { }.
Prompt:
"Write content for a PrestaShop product using the data below. Do not invent any technical specs; only use values provided. Output must be valid CSV (UTF-8) with header row matching PrestaShop import columns: 'Active, Name [en], Categories, Price tax excluded, Price tax included, Quantity, Reference, Short description [en], Description [en], Tags [en], Meta title [en], Meta keywords [en], Meta description [en], URL rewritten [en], Image URLs'. Wrap 'Short description' and 'Description' in simple HTML (
,
, - , ) where needed. Keep Meta title ≤ 60 characters and Meta description ≤ 160 characters. Provide an SEO-friendly URL slug (lowercase, hyphens, no accents). Provide alt text for each image (comma-separated after Image URLs, format: url|alt). Use this input: {product_name}, {brand}, {category_list}, {price_excl}, {price_incl}, {quantity}, {reference}, {features_list}, {materials}, {target_audience}, {tone: e.g., friendly/professional}, {language}. If any field is unknown, leave blank."
Example to paste with values:
{product_name: "Nordic Wool Throw Blanket 130x180 cm", brand: "ScandiHome", category_list: "Home & Living > Bedding", price_excl: "49.00", price_incl: "58.80", quantity: "120", reference: "SH-BLANKET-130", features_list: "100% wool; machine washable; lightweight but warm", materials: "Mulesing-free wool", target_audience: "urban homeowners, 25-55, s-conscious", tone: "warm, stylish", language: "en", image_urls: "https://example.com/images/blanket1.jpg,https://example.com/images/blanket2.jpg"}
Product output expectations:
- CSV row matching header
- Short description: 1–3 short HTML paragraphs or 3 bullet points (max 300 chars)
- Description: longer HTML product page copy (150–350 words), features in bullets, washing/care, one CTA
- Meta title: ≤60 chars, include primary keyword
- Meta description: ≤160 chars, include primary keyword and CTA
- URL slug: lowercase hyphenated
- Tags: comma-separated keywords
- Image alt texts appended to Image URLs as url|alt
Example short sample prompt you can paste directly:
"Create a PrestaShop product CSV for the product below. Output only CSV with header described earlier. Use the language 'en'. Do not invent specs. Input: name='Nordic Wool Throw Blanket 130x180 cm', brand='ScandiHome', categories='Home & Living > Bedding', price_excl='49.00', price_incl='58.80', quantity='120', reference='SH-BLANKET-130', features='100% wool; machine washable; lightweight', materials='Mulesing-free wool', target_audience='urban homeowners 25-55', tone='warm stylish', image_urls='https://example.com/images/blanket1.jpg,https://example.com/images/blanket2.jpg'"
Bulk product generation tips
- Provide a CSV/JSON with your base columns (name, reference, price, features) and ask the model: "For each row, generate the following fields: Meta title, Meta description, Short description (HTML), Long description (HTML), URL slug, Tags, image alt texts. Return a CSV with original columns plus generated ones."
- Batch in chunks (e.g., 50–200 items) to avoid rate/size issues.
- Validate automatically: check meta length, unique URL slugs, no invented specs.
Category prompt template
For category pages you want SEO-friendly copy and navigation suggestions.
Prompt:
"Create SEO-optimized category content for PrestaShop. Output JSON with keys: name, meta_title(≤60), meta_description(≤160), url_slug, short_description_html (80–160 words), long_description_html (150–400 words), suggested_filters (comma-separated), hero_h1, image_alt_text. Tone: {tone}. Target audience: {audience}. Primary keywords: {keywords}. Do not invent product data; focus on category benefits, typical product types, and buying signals."
Category examples: seasonal category copy, gift guides, "Outlet" vs "New Arrivals" differences.
CMS page prompt templates (About Us, Shipping, Returns, Terms, FAQs)
General template:
"Write a {page_type} CMS page suitable for PrestaShop. Output only HTML for the page body plus a line with meta_title (≤60), meta_description (≤160), and URL slug. Tone: {tone}. Important facts to include: {business_facts}. Legal/compliance points: {compliance_requirements}. Provide a short 'What this means for customers' section and 5 FAQs."
Examples:
- About Us: brand story, mission, values, sustainability claims (only if true), CTA (newsletter/signup)
- Shipping: processing time, shipping partners, estimated delivery, returns link, CTA to shipping calculator
- Returns/Refunds: clear steps, deadlines, exceptions, contact method, overview + bulleted steps
Prompts for translations and localization
- Ask for cultural localization, measurements conversion (cm ↔ inches), currency conversion, local idioms.
- Template:
"Translate and localize the following product description into {target_language}. Convert measurements to {units} and currency to {currency_symbol}. Adjust phrasing to {region} audience and preserve SEO keywords: {keywords}. Return only the translated meta_title, meta_description, short_description_html, description_html, url_slug."
Prompts for images, alt text, and file names
- Generate descriptive alt text that includes the main keyword and key attributes (color, model, angle).
Prompt:
"For each image URL, produce an alt text (max 125 characters) describing the product, color, and context. Return CSV where each row is 'image_url|alt_text'."
Prompts for attributes and combinations (size/color)
- Supply a list of attributes and ask for canonical attribute labels and combinations.
Prompt:
"Given attributes: sizes S,M,L,XL; colors: red, navy, black; produce PrestaShop-ready combinations table as CSV with columns: 'Reference, Attribute (size), Attribute (color), EAN13 (leave blank if none), Qty, Price impact'. Base reference: {ref}, base_qty: {qty}."
SEO/keywords guidance
- Provide primary keyword and 4–6 LSI/secondary keywords.
- Instruct the model to place the primary keyword in meta title, first sentence, one H2, and within 2–3 times in the long description (natural).
- Avoid keyword stuffing. Keep readability first.
Quality control prompts (review and editing)
- Prompt to audit content:
"Review the following product descriptions for consistency with brand voice, detect hallucinated specs, check SEO (meta lengths, keyword presence), and suggest edits. Output a JSON with fields: issues_found[], suggested_edits_html, confidence_score (1-10)."
- Also ask model to generate A/B variants for testing.
Practical ready-to-use prompt examples
1) Single product (short):
"Write a short and long product description for 'Zora Ceramic Coffee Mug, 350ml, matte black' for PrestaShop. Tone: modern, minimal. Short description: max 300 chars, HTML. Long description: 120–250 words, include 5 bullet feature list and one CTA. Create Meta title ≤60 chars, Meta description ≤160 chars, and URL slug. Primary keyword: 'ceramic coffee mug'. Do not invent materials beyond 'ceramic' or dimensions beyond '350ml'. Return JSON: {name, short_description_html, description_html, meta_title, meta_description, url_slug}."
2) Bulk CSV (multi-row)
"Using the attached CSV rows with columns {name, brand, price_excl, price_incl, quantity, reference, features, images}, generate for each row: meta_title (≤60), meta_description (≤160), short_description_html, description_html, url_slug, tags, image_alt_texts. Return a CSV with original columns plus these new columns. Do not invent specs."
Safety & verification reminders
- Always verify technical specs and compliance claims (e.g., "organic", "waterproof", certifications).
- Run automatic checks for duplicate slugs and meta lengths.
- Human-review sample outputs before publishing at scale.
- For legal/terms pages consult a lawyer before publishing.
Automation tips
- Use consistent placeholders and a fixed output schema so you can parse results reliably.
- Use the model to also produce QA checks and CSV validation steps.
- Keep batches small enough for context limits; store product base data and only request generation per chunk.
If you want, I can:
- Convert one of the templates into a ready-to-run prompt for ChatGPT with your store’s exact fields, brand voice and a sample product list.
- Generate a short demo CSV for 3 example products so you can see exact output structure.
Which would you like: a ready-to-run prompt prefilled for your store, or a 3-product demo CSV output?
