Below is a compact, practical guide + ready-to-use prompt templates you can paste into ChatGPT or any LLM to create high-quality content for PrestaShop products, categories and CMS pages. It covers structure, required fields, SEO/formatting rules, bulk generation, and validation tips.
Quick principles
- Be explicit about role, audience, voice, and output format (JSON/CSV/HTML/plain).
- Provide the exact PrestaShop fields you need filled and examples of values.
- Ask for length limits (meta title <60 chars, meta description ~120–155, short desc ~50–160).
- Require machine-readable output (CSV/JSON) when doing bulk work so it's import-ready.
- Ask the model to flag missing data and to not invent facts (or only invent plausible filler when allowed).
- Provide brand s (tone, key selling points, required keywords) and locale/language.
Common PrestaShop fields to include
- Product: name, short_description, description (HTML), meta_title, meta_description, meta_keywords, friendly_url, tags, price, reference, sku, ean13, upc, quantity, weight, condition, manufacturer, supplier, visibility, available_for_order, online_only, show_price, features (name:value), attributes/combinations, images (filenames), attachments, customizations.
- Category: name, description (HTML), meta_title, meta_description, friendly_url, parent_category, thumbnail, cover_image.
- CMS page: meta_title, meta_description, friendly_url, content (HTML), active.
“Perfect” general prompt skeleton
You can reuse this skeleton and fill the brackets.
You are a senior e-commerce copywriter experienced with PrestaShop. Write content for [PRODUCT|CATEGORY|CMS] page(s) in [LANGUAGE], targeted at [AUDIENCE]. Tone: [TONE]. Brand guidelines: [short bullet points]. SEO keywords to prioritize: [keywords]. Output must exactly follow [JSON/CSV] schema below. Do not include extra commentary.
Schema: [insert schema; e.g. JSON with fields]. Constraints:
- meta_title ≤ 60 characters
- meta_description 120–155 characters
- friendly_url: lowercase words separated by hyphens, no special chars
- short_description: plain text ≤ 160 chars
- description: clean HTML with
,
, - , only
- If data missing, return null for that field and add "missing_fields" array
Now produce [one item / N items] using this schema. Example input data: [paste table or JSON rows]. Begin output with exactly: “BEGIN_OUTPUT” and end with “END_OUTPUT”.
Example prompts (ready-to-use)
1) Single product — JSON output
You are a senior ecommerce copywriter. Create product content for PrestaShop using the JSON schema below. Write in English, tone: helpful and persuasive. Use the provided product data; do not invent specs (except fill missing marketing copy as plausible). Prioritize SEO keywords: "wireless headphones", "noise cancelling", "long battery life". Constraints: meta_title ≤ 60 chars, meta_description 120–155 chars, friendly_url lowercase hyphen-separated, short_description ≤ 160 chars. Description should be HTML paragraphs and bullets.
Schema (output JSON array of one object):
{
"name": string,
"short_description": string,
"description": string, // HTML allowed (
,
, - , )
"meta_title": string,
"meta_description": string,
"meta_keywords": string,
"friendly_url": string,
"tags": [string],
"features": [{"name": string, "value": string}],
"images": [string], // filenames
"missing_fields": [string] // fields you couldn't fill
}
Input data:
{
"name": "AeroSound X2 Wireless Headphones",
"manufacturer": "AeroSound",
"battery_life": "45 hours",
"noise_cancellation": "Active noise cancellation (ANC)",
"weight": "240 g",
"colors": "Black, Silver"
}
Return only the JSON output between BEGIN_OUTPUT and END_OUTPUT markers.
2) Bulk CSV — many products for import
You have a CSV with columns: id, name, short_description, description, meta_title, meta_description, meta_keywords, friendly_url, tags, features, images, language. For each row produce completed CSV row. Preserve commas using proper quoting. Do not output any extraneous text—only CSV. Keep meta lengths as specified. If product data is insufficient, populate marketing fields generically and add “[MISSING: fieldname]” into a notes column.
3) Category creation
You are an SEO copywriter creating PrestaShop category content. Language: French. Tone: professional. For each category record (name, parent_category, key_features, target_audience) produce: name, description (HTML paragraphs and bullet list), meta_title (≤60), meta_description (120–155), friendly_url. Use keywords: [list]. Return JSON array.
4) CMS page (policy/about)
Create an English CMS page "Shipping & Returns" for an ecommerce store. Tone: clear, friendly, and legally safe. Sections: summary, shipping options & times, costs, international shipping, returns & refunds policy, contact. Output as HTML content string plus meta fields. Provide a one-line “last updated” note and mention the effective date.
5) Product combinations / attributes
Given base product colors and sizes, output a list of combinations as CSV with columns: combination_reference, attributes (e.g., Color:Black;Size:M), ean13, price_impact, quantity, default_on. Use consistent reference format PREFIX-Color-Size.
6) Image alt text generator
For each product image filename, produce descriptive, SEO-friendly alt text no longer than 125 characters. Include product name and relevant feature.
Examples filled (short)
Example — single product result (trimmed)
BEGIN_OUTPUT
{
"name": "AeroSound X2 Wireless Headphones",
"short_description": "Comfortable wireless headphones with ANC and up to 45 hours battery life.",
"description": "
The AeroSound X2 delivers premium audio with active noise cancellation and a long-lasting 45‑hour battery. Perfect for commuting and travel.
- Active Noise Cancellation
- 45 hours playback
- Lightweight 240 g design
",
"meta_title": "AeroSound X2 — ANC Wireless Headphones",
"meta_description": "AeroSound X2 wireless headphones with active noise cancellation and 45 hours battery life. Comfortable, lightweight and travel-ready.",
"meta_keywords": "wireless headphones, ANC, long battery life, AeroSound",
"friendly_url": "aerosound-x2-wireless-headphones",
"tags": ["headphones","wireless","ANC"],
"features": [{"name":"Battery life","value":"45 hours"},{"name":"Weight","value":"240 g"}],
"images": ["aerosound-x2-1.jpg","aerosound-x2-2.jpg"],
"missing_fields": []
}
END_OUTPUT
Practical tips & rules
- Provide seed data: always include at least name, core specs, target keywords, and language.
- Use “do not invent” vs “fill plausibly” flags depending on how strict you need accuracy.
- Ask for machine-checked constraints: tell the model to count characters for meta fields.
- For CSV/Excel import, require exact column order and that the model escape commas and quotes properly.
- For HTML description, ask to limit tags to safe set (,
, - , , ).
- For translations, ask for both original and translated content in separate fields, or provide existing descriptions and ask to rewrite for SEO+locale.
- For A/B tests: ask for two meta titles and two meta descriptions.
Validation checklist (after generation)
- Meta title ≤ 60 chars; meta description 120–155 chars.
- Friendly URL: lowercase, no spaces, hyphens between words.
- No invented technical specs unless allowed.
- HTML has only allowed tags and is well-formed.
- No forbidden words or claims (e.g., medical claims).
- Run spellcheck and locale-specific grammar check.
Short example prompts you can copy-paste
- Product (short): "You are an ecommerce copywriter. Given product: name='Classic Leather Tote', features='100% leather; 2 interior pockets; 12 x 14 in', audience='professional women', SEO keywords='leather tote bag, work tote'. Output JSON with fields: name, short_description (≤160 chars), description (HTML), meta_title (≤60), meta_description (120–155), friendly_url. Do not invent materials or sizes. Begin with BEGIN_OUTPUT and end with END_OUTPUT."
- Bulk CSV (short): "I will paste a CSV of products. For each row, generate completed meta_title, meta_description and friendly_url columns. Return only CSV with the same columns + meta_title, meta_description, friendly_url."
Final notes
- Start with a small test batch and verify import behavior in PrestaShop before full-scale generation.
- Use the schema approach to avoid hallucinations and ensure direct importability.
- If you want, paste one product or a CSV sample and I’ll create the exact prompt (and a generated output example) tailored to your data.
Would you like a ready-made prompt tailored to a real product or a CSV sample you have?
