On SEO forums and Reddit threads, the same rumour keeps coming back: “Google killed FAQ Schema in 2023”. The reality is more nuanced. In August 2023, Google did drastically reduce the display of FAQ rich snippets — only “authoritative government and health sites” kept standard display. But on e-commerce product pages, FAQ Schema continues to work and produce visible results in the SERP in 2026.
On the PrestaShop 8 stores we audit, FAQ Schema properly implemented on product pages still produces:
- visible snippet extensions in the SERP (questions appearing below the result title);
- better product understanding by Answer Engines (ChatGPT, Perplexity, Google AI Overviews) which consume this structured data;
- an editorial quality signal that contributes to the overall authority of the product page.
The problem is that FAQ Schema is also one of the most misunderstood markups: implementations that don’t validate, JSON-LD that validates but that Google never displays, FAQs written for the markup rather than for buyers. This guide covers the correct implementation on PrestaShop 8, validation, and — most importantly, the part everyone gets wrong — debugging when your FAQ Schema is valid but doesn’t appear in Google results.
Why FAQ Schema is still worth it in 2026
The “FAQ Schema is dead” rumour stems from a Google update in August 2023 that did limit FAQ rich snippet display. But the limitation applies to editorial pages (blog articles, help pages) — not to e-commerce product pages.
On product pages, FAQ Schema remains active for three reasons:
- Google treats product FAQs as commercial structured data in the same way as Product Schema or Review Schema. This data feeds both the classic SERP and enriched features (Shopping, Discover, AI Overviews).
- Answer Engines (ChatGPT, Perplexity, Claude, Google AI Overviews) have heavily consumed this data since 2024. When a user asks ChatGPT “what’s the difference between product X and Y”, the AI regularly cites structured FAQs found on product pages — often with a source link. This is AEO (Answer Engine Optimization) applied to e-commerce.
- FAQ Schema remains an editorial authority signal that Google interprets as “this page was designed to answer buyers’ questions”, beyond simple SERP rendering.
On stores we monitor, product pages with well-implemented FAQ Schema show organic CTRs 12 to 25% higher than pages without (measured on cohorts equivalent in terms of positions and keywords). The return is more moderate than in 2022 when display was systematic, but remains clearly positive relative to the implementation cost.
What Google expects exactly: valid JSON-LD structure
Google recognises three structured data formats: JSON-LD, Microdata, and RDFa. JSON-LD is officially recommended by Google and is the only format you should use in 2026.
JSON-LD is placed in a script tag of type application/ld+json in the head or body of the page (both are valid; placing it in the head is cleaner). Here is the minimum structure of a valid FAQPage Schema for a product page:
{
"@context": "https://schema.org",
"@type": "FAQPage",
"mainEntity": [
{
"@type": "Question",
"name": "What size do you recommend?",
"acceptedAnswer": {
"@type": "Answer",
"text": "Please refer to our size guide in the dedicated section. Most customers take their usual size; for fitted models, go one size up."
}
},
{
"@type": "Question",
"name": "What is the delivery time?",
"acceptedAnswer": {
"@type": "Answer",
"text": "Delivery within 48h in mainland France, 3-5 days in Europe, 7-10 days to the rest of the world."
}
}
]
}A few rules that determine success or failure:
The content must visibly exist on the page. Google only takes FAQ Schema into account if the questions and answers are actually displayed to the visitor (in an accordion, collapsible block, whatever). “Hidden” FAQs that only exist in the JSON-LD are ignored and can even trigger a manual penalty for cloaking.
No promotion, no marketing in the answers. Google has explicitly banned FAQ Schemas containing promotional content or external links. Answers must be informational, neutral, and useful to the buyer.
Only one FAQPage per URL. If you have multiple FAQ sections on a page (e.g. a general FAQ plus a shipping FAQ), group everything into a single mainEntity array. Do not create multiple separate FAQPage JSON-LD blocks.
At least 2 questions per page. A single question + answer is not a FAQPage, it’s an isolated Q&A. Aim for 4 to 8 questions per product page as the sweet spot — enough to be informative, not so many as to become unreadable.
Encoding special characters. Apostrophes, quotes, and special characters must be properly escaped. A misplaced comma in the JSON breaks the markup and the Google validator rejects everything.
Implementation on PrestaShop 8: 3 methods from simplest to most powerful
PrestaShop 8 does not provide native FAQ Schema on product pages. Here are the three possible approaches.
Method 1 — Manual in the Twig template
The fastest approach for testing: edit the template themes/your-theme/templates/catalog/product.tpl and add the JSON-LD at the end of the file.
This method works for stores with a small catalogue (fewer than 50 products) where FAQs can be managed via a custom feature (feature_value_lang.value containing the formatted FAQ text). The template reads the feature and generates the JSON-LD.
Obvious limitations: FAQs are not editable from the product back office on a product-by-product basis with a dedicated editor — it becomes copy-paste in a text feature that quickly becomes unmanageable. For 5 products, it’s workable. For 500, it’s inhumane.
Method 2 — Module with per-product FAQ
The standard approach: a module that adds a “FAQ” tab to the product back office, where you edit questions and answers like a mini-CMS. The module then injects the JSON-LD into the page head (via the displayHeader hook) and displays the visible FAQs in an accordion at the bottom of the product page.
This method solves the editorial problem but introduces the next one: the FAQs have to be written. On 500 products, that’s several days of manual writing. Most merchants give up after 30 products, and the module becomes a graveyard of empty FAQs.
Method 3 — AI-automated generation
The modern approach: a module that automatically generates FAQs per product using AI (OpenAI or Claude), based on the product name, description, characteristics, and reviews. You trigger bulk generation from the product list, review and adjust the generated FAQs, and the module pushes the JSON-LD into the head.
This is what our DataFirefly FAQ AI Product module does: bulk generation, manual validation, FAQs visible on the front end with valid JSON-LD in the background, all in multi-shop and multilingual mode.
Practical recommendation: start with method 1 on 5 product pages to validate that Google properly indexes your markup, then switch to method 3 as soon as you want to scale beyond 30 products. Method 2 is rarely the right compromise — either you want to scale (method 3), or you’re managing a handful of premium pages manually (method 1).
Validating your FAQ Schema before and after going live
Validation happens in two steps: syntactic validation (is the JSON-LD well-formed?) and Google validation (does Google recognise the structure?).
Syntactic validation. First, paste your JSON-LD into a standard JSON validator (jsonlint.com for example). If the structure is syntactically broken (missing comma, orphaned bracket), the Google validator will ignore everything. This step catches 90% of problems in 30 seconds.
Google’s Rich Results Test. The official tool at search.google.com/test/rich-results. Paste your product page URL (or directly the HTML source code), and the tool tells you:
- whether the FAQ Schema is detected;
- how many questions are recognised;
- if there are errors or warnings;
- what the rich snippet would look like (preview).
A valid FAQ Schema displays “FAQPage” in the list of detected items with a count of questions.
Search Console validation. Once the product page has been indexed by Google, go to Search Console > Enhancements > FAQ. If Google has accepted the markup, the page appears in the “Valid pages” report. If there are issues, it appears in “Pages with warnings” or “Pages with errors”.
The indexing delay. The markup you publish today won’t appear in Search Console for 2 to 7 days on an existing product page. On a new page, allow 1 to 3 weeks before Google crawls, indexes, and recognises the markup. During this delay, the Rich Results Test remains your best validation tool.
The classic trap: your FAQ Schema is valid but doesn’t appear in Google
This is the most common search query from frustrated e-commerce developers: “prestashop faq schema not showing in google”. The markup validates, the Rich Results Test passes, Search Console reports nothing — yet no rich snippet appears in the SERP. Here are the causes in order of frequency.
Cause 1: Google only displays the rich snippet for certain queries, not systematically. Since the August 2023 update, Google only triggers the FAQ rich snippet for queries where it judges FAQs to provide genuine informational value to the searcher. A highly transactional query (“buy red shoes size 42”) will probably never show the FAQ snippet, even if the product page has perfect markup. A more informational query (“how to choose running shoe size”) may trigger display. Conclusion: the markup exists and is valid, but Google decides when to show it.
Cause 2: the product page isn’t ranked highly enough to trigger rich features. Google only enriches SERPs above a certain page quality threshold. If your page is on page 2 for the query, the FAQ snippet will almost never appear — Google reserves enrichment for organic results at the top of page 1.
Cause 3: FAQs are duplicated across multiple pages. If your 50 product pages all display the same 4 FAQ questions (e.g. generic FAQs about shipping, returns, customer service), Google detects the duplication and ignores the markup on duplicated pages. This is a common cause when generic FAQs are added without customising per product.
Cause 4: the markup isn’t crawled (robots.txt or JS-only). Check that the JSON-LD appears in the raw HTML served by your server (curl or Developer Tools > View page source). If the markup is generated in JavaScript after page load, Google may not execute it on some pages. Server-side rendering is required.
Cause 5: manual action or algorithmic penalty. Rare but worth checking in Search Console > Security & manual actions. If Google has manually penalised your site for structured data spam (overly promotional FAQs, irrelevant answers), all your rich snippets are cut off.
Cause 6: FAQs are written for the markup, not for buyers. This is the subtle trap. If your FAQs are artificial Q&As designed solely to fill the JSON-LD, they don’t trigger rich snippets because Google identifies that they provide no real informational value. Symptom: valid markup, FAQ rich snippet never displayed, no CTR improvement despite weeks of waiting.
The test: show your FAQs to someone who doesn’t know your catalogue. If the questions seem obviously artificial (“why is this product great?”), it’s a dead end. If they look like questions a buyer would sincerely ask (“is this model suitable for this use case?”), you’re on the right track.
Why AI beats manual copy-paste at scale
Manual FAQ writing remains viable up to 30–50 product pages. Beyond that, the writing hours become unreasonable, and quality drops because the writer gets tired and produces generic FAQs to finish the batch.
Well-used AI solves both problems:
Contextual generation. Good AI takes as input the product name, description, characteristics, technical specifications, and customer reviews. It produces FAQs that reflect the real questions buyers ask (since it has learned from millions of product pages and billions of user queries). This is significantly more relevant than generic template FAQs.
Multilingual consistency. On a multilingual store, AI generates FAQs in all active languages in one pass, with translation quality generally superior to machine translation applied to FAQs written in one language then translated.
Speed. On 500 product pages, AI generation takes a few minutes (in parallel, in batches). Manual validation of generated FAQs then takes 10 to 20 seconds per page (read and adjust if needed). Total: 2 to 4 hours to process 500 products, versus 20 to 40 hours of manual writing.
The risk to know: poorly guided AI can produce generic FAQs (“what are the benefits of this product?”) that add nothing. Result quality depends on the prompt and model. Our DataFirefly FAQ AI Product module uses GPT-4 or Claude as options, with prompts specifically designed to produce concrete, buyer-useful questions — not tautological questions. Manual validation remains possible and recommended for strategic pages.
Beyond FAQ Schema: Product, Review, Breadcrumb for stacking rich snippets
FAQ Schema alone produces a moderate effect. The real SEO gain on product pages comes from stacking multiple structured markups.
Product Schema. The baseline: name, price, currency, availability, image, brand, GTIN or EAN if you have it. Without Product Schema, enriched rich snippets (price in SERP, stock label, visible breadcrumb) never appear. PrestaShop 8 natively generates a basic but incomplete Product Schema — verify it includes priceCurrency, availability at schema.org/InStock or equivalent, and image with an absolute URL.
Review and AggregateRating Schema. If your product page has verified customer reviews, the aggregateRating (average rating out of 5, number of reviews) triggers stars in the SERP — one of the most powerful CTR levers in e-commerce. Our DataFirefly Verified Reviews module generates this markup automatically from collected reviews and injects it into the existing Product Schema.
BreadcrumbList Schema. To show the breadcrumb in the SERP instead of the raw URL. Positive visual effect but modest impact on CTR. PrestaShop 8 generates this markup natively on category and product pages; verify it hasn’t been broken by your custom theme.
Offer Schema. If you have promotions or struck-through prices, Offer markup allows displaying the reduced price with the original price in the SERP. Even more powerful during sales and Black Friday. Compatible with Product Schema by nesting it in offers.
Cross-markup consistency. All schemas on the same page must be consistent: if your Product Schema says “price €49” and your Offer Schema says “price €39”, Google considers the markup suspicious and may ignore everything. Verify that values propagate correctly.
Conclusion: a moderate investment that remains profitable in 2026
FAQ Schema on product pages is no longer the hyper-profitable lever of 2022. But it remains one of the most profitable structured markups in 2026, particularly when combined with other schemas (Product, Review, Breadcrumb) to form a coherent stack.
Manual implementation is viable up to 30–50 pages. Beyond that, switch to AI generation — 90% time saving and superior quality because you cover 100% of the catalogue instead of a handful of pages.
For related SEO topics, browse our E-commerce SEO category, or our PrestaShop tutorials for technical topics. And if you want to skip the manual writing phase and go straight to scale, the DataFirefly FAQ AI Product module generates your FAQs and their JSON-LD markup for your entire PrestaShop 8 catalogue in minutes.