PrestaShop SEO: friendly URLs, canonical redirects, robots.txt, schema and speed

PrestaShop technical SEO comes down to four screens: SEO & URLs (friendly URLs, the canonical redirect and robots.txt), Product Settings (what happens to disabled products), the Google sitemap module and Performance (caching and CCC). This guide walks through each setting in PrestaShop 9.2, the current release, with a tested module for AI crawlers.

To check it on your own site: Online shops

PrestaShop is an open-source ecommerce platform written in PHP. It controls URLs, canonicals, robots.txt, the sitemap (through a module) and the theme's product markup; it does not control compression, headers or the PHP version on your server. Version 9.2.0 was published on 30 September 20261 and runs on PHP 8.1 to 8.52. Since 9.1, new installations ship with the Hummingbird theme instead of Classic3. What applies to every online shop is in the guide to ecommerce SEO for Google and AI search.

Where is each SEO setting in PrestaShop 9?

How do you enable friendly URLs and the canonical redirect?

In SEO & URLs, under Set up URLs4:

1. Friendly URL: turn it on if your server supports URL rewriting (on Apache, mod_rewrite). PrestaShop warns that if the server does not support it, the store can become unavailable.
2. Accented URL: allows non-ASCII characters in URLs. Switched off, URLs carry no accents, which makes them easier to copy and link to.
3. Redirect to the canonical URL: decides how the store responds when someone requests a page through a URL that is not its own. There are three options: no redirection, 302 and 301.

In PrestaShop 9.2 the installation sets this last option to 302, and the admin itself recommends «301 Moved Permanently» once the store has gone live5. Change it. A 302 tells Google that the right URL might change.

Schema of URLs is where you define routes. In the 9.2 code, the product route is {id}{-:id_product_attribute}-{rewrite}.html, which produces /12-t-shirt.html or, with a combination, /12-45-t-shirt.html6. The number cannot be removed with the default routes4. A product page's canonical points to the product URL without a combination7, so combinations do not compete with each other. Change the format once the store is indexed and every URL changes: decide before launch.

What does PrestaShop do with filters and pagination?

The faceted search module adds filters to the category URL with the q parameter (for example, ?q=Size-M), and sorting uses order8. A category's canonical keeps only the page number and drops the filters9, which is what Google asks for: every page in a series crawlable and canonical to itself10, and filters out of the index11.

The robots.txt PrestaShop generates fits that. It blocks ?q=, ?order=, ?tag=, ?id_currency=, ?search_query=, ?back= and ?n= (after & too), the cart, account, search and order controllers, and internal folders such as classes/, config/ and vendor/; and it allows module CSS, JS and images12. It does not block ?page=. We checked it with a matcher that applies Google's rules: /3-clothes and /3-clothes?page=2 stay allowed; ?q=Size-M, ?order=product.price.asc and ?page=2&q=Colour-Blue are blocked.

If a filter combination has its own search demand, do not open up the filter: create a category with its own copy and URL.

How do you add rules for AI crawlers without losing them?

The Generate robots.txt file button replaces the whole file, and the documentation says your own rules must be added after PrestaShop has generated its version4. The problem is the next generation: it wipes what you added. Generation fires two hooks, actionAdminMetaBeforeWriteRobotsFile and actionAdminMetaAfterWriteRobotsFile12, so a small module can write your block every time. Save it as modules/robotsia/robotsia.php, install it and regenerate robots.txt:

<?php
if (!defined('_PS_VERSION_')) {
    exit;
}

class RobotsIa extends Module
{
    public function __construct()
    {
        $this->name = 'robotsia';
        $this->tab = 'seo';
        $this->version = '1.0.0';
        $this->author = 'Your store';
        $this->ps_versions_compliancy = ['min' => '8.0.0', 'max' => _PS_VERSION_];
        parent::__construct();
        $this->displayName = 'robots.txt: AI crawlers';
        $this->description = 'Adds a group for training crawlers every time robots.txt is generated.';
    }

    public function install()
    {
        return parent::install() && $this->registerHook('actionAdminMetaAfterWriteRobotsFile');
    }

    public function hookActionAdminMetaAfterWriteRobotsFile(array $params)
    {
        $fd = $params['write_fd'];
        fwrite($fd, "\n# Model training: blocking these does not remove you from AI answers\n");
        foreach (['GPTBot', 'ClaudeBot', 'Google-Extended', 'CCBot'] as $bot) {
            fwrite($fd, 'User-agent: ' . $bot . PHP_EOL);
        }
        fwrite($fd, 'Disallow: /' . PHP_EOL);
    }
}

How we tested it: php -l in the official php:8.3-cli image, and a run of the hook with the same write order as Tools::generateRobotsFile in 9.2: the group appears at the end, after PrestaShop's rules. It leaves the User-agent: * group alone, so OAI-SearchBot, Claude-SearchBot and PerplexityBot (the crawlers that decide whether ChatGPT, Claude or Perplexity can cite you) follow the general rules. GPTBot, ClaudeBot and CCBot collect content to train models, and Google-Extended does not affect Google Search1314. The full list is in the guide to AI crawlers and robots.txt.

How is the sitemap generated?

PrestaShop generates it with the Google sitemap module (gsitemap), which creates one sitemap per language and stays current through a cron URL shown in its configuration15. Order matters: robots.txt only includes the Sitemap: line if the sitemap index file (1_index_sitemap.xml for shop 1) already exists when you generate it12. Generate the sitemap first, then robots.txt.

The module ships with priorities per page type, but do not spend time on them: Google ignores <priority> and <changefreq>16.

What happens when you disable a product?

Under Shop Parameters > Product Settings, the default behaviour for disabled products is an error page with a 404 response; the other options are a 410, or showing the product as discontinued with a 200, 404 or 410 response17. On each product, the SEO tab has a redirection setting for when the product is offline, which allows a permanent (301) or temporary (302) redirect to another product or to a category18.

The rule of thumb: if the product is not coming back and has a clear replacement, 301 to the replacement or its category; if it is not coming back and there is no replacement, 410; if it is coming back soon, show it as discontinued with a 200 so you keep the URL.

What structured data does the theme output?

Classic and Hummingbird generate a product JSON-LD block in templates/_partials/microdata/product-jsonld.tpl1920. This is what Hummingbird 2.1.2 prints, against what Google asks for in merchant listings21:

Two details. The markup's description comes from the page's meta description, not from the product description: fill it in on every product. And if you assign no manufacturer, the brand becomes your store's name; for third-party products, assign the real manufacturer. For shipping and returns, Google recommends declaring the policies once in your Organization markup21. More detail in the guide to product schema.

How do you force HTTPS and speed up the store?

Under Shop Parameters > General, turn on Enable SSL. In the 9.2 code, with SSL on, the store serves every page over https and 301-redirects requests that arrive over http22. HSTS and the other headers belong on the server: the guide to Apache HTTPS, HSTS and security headers has the snippets.

Under Advanced Parameters > Performance23:

1. Smarty: in production, keep caching on and choose never to recompile template files; Force compilation is for development only.
2. CCC: Smart cache for CSS and for JavaScript are on in a fresh installation24. PrestaShop asks you to test the store with the JavaScript one, because combining scripts can sometimes cause problems.
3. Debug mode: off in production.

Measure the result against Google's thresholds at the 75th percentile: LCP of 2.5 s or less, INP of 200 ms or less and CLS of 0.1 or less25.

How to check it

• Canonical redirect: curl -I https://www.example.com/12-wrong-name.html should return a 301, not a 302.
• robots.txt: open /robots.txt, look for your block at the end and for the Sitemap: line, and review the robots.txt report in Search Console.
• Canonical: Search Console URL Inspection on a product with a combination and on a filtered category.
• Markup: run a product page through the Rich Results Test.

What SmoothSeen does with this

SmoothSeen analyses a product or category page and checks what this guide configures: robots.txt, sitemap, canonical and noindex, product structured data, speed with real-user Core Web Vitals (or Lighthouse when there is no field data), compression, HTTPS and headers. On the AI side it checks which crawlers your robots.txt allows, keeping search and training apart, and whether your server returns a 403 to any of them.

What to do this week

Open SEO & URLs and check that Redirect to the canonical URL is set to 301. Then generate the sitemap, regenerate robots.txt and confirm it carries the Sitemap: line. To see the rest of the list measured, analyse your online store with SmoothSeen.

Frequently asked questions

Can you remove the number from PrestaShop URLs?

Not with the default routes: the numeric ID is part of the product and category routes, and PrestaShop uses it to find the page. Third-party modules can remove it, but they change every URL in the store. It only pays off on a new store or in a planned migration with 301 redirects from every old URL.

Do you need to regenerate robots.txt after updating PrestaShop?

It is worth checking. The file does not update itself: it reflects the rules of the version you generated it with and the languages active at that time. After a major update, or when you add a language, generate it again from SEO & URLs. If you added rules by hand, move them into a module first so you do not lose them.

Does PrestaShop block ChatGPT or Perplexity?

No. The generated robots.txt only has the User-agent: * group, with blocks on parameters, private controllers and internal folders. OAI-SearchBot and PerplexityBot can read product and category pages. If you are not being cited, check first whether your server or firewall answers them with a 403 while serving the page to a browser.

Are PrestaShop's meta keywords worth filling in?

Not for Google. PrestaShop still has a keywords field on some pages, but Google documents that the meta keywords tag has no effect at all on indexing or ranking in Google Search. Spend that time on the meta title, the meta description and the product description, which do show up in results.