Drupal SEO with Snoika
Drupal SEO with Snoika adds new pages built for Google and AI assistants to your Drupal site. Snoika connects at the DNS or CDN level and publishes pages on your own domain, while Drupal and its SEO modules keep running as they are. To start, connect your domain through the CDN you already use and approve your first pages.
- Cloudflare Worker, which you set up yourself with…
- Fastly, which Snoika's team sets up
- Google Cloud Load Balancer
Trusted by
Connecting Drupal to Snoika
You connect Drupal to Snoika in your CDN or web server. Snoika sits in front of the whole domain. When a request is for a page Snoika manages, Snoika serves it. Every other request goes through to Drupal unchanged. The connection covers the entire domain, so Snoika can add a new page or update an existing one anywhere on the site, and no fixed path is set aside for it.
What you need before you start
You need access to one of these setups:
- Cloudflare Worker, which you set up yourself with Snoika's guide
- Fastly, which Snoika's team sets up
- Google Cloud Load Balancer
- AWS CloudFront
- Nginx
- A setup tailored to your stack, on request
Steps to turn it on
- Pick the setup that matches where your DNS or CDN already lives.
- Add the routing yourself with the Cloudflare guide, or work through it with Snoika's team for the other setups.
- Snoika builds page templates that match your Drupal theme, so its pages look like the rest of the site.
- Snoika configures Google Analytics 4 (GA4) and Google Tag Manager events, so you can track the leads each page brings in.
- Read each draft in the Snoika dashboard and approve it before it goes live.
- Schedule the pages you've approved. For page types that need no review, schedule generation with automatic publishing instead.
What Drupal modules still handle
You still need your Drupal modules for every page Drupal renders. On most sites that means Metatag alongside Pathauto and Redirect. A sitemap module such as Simple XML Sitemap rounds out the setup. Snoika never touches Drupal's database or code, so those modules carry on as before. The table shows where each job sits once both systems are running.
| Job | Drupal modules | Snoika |
|---|---|---|
| Meta tags on existing nodes | Metatag | Not involved |
| URL aliases and redirects | Pathauto, Redirect | Not involved |
| New researched pages | Your editors write them in Drupal | Created and published in 135+ languages |
| Schema, hreflang and internal links | Schema.org Metatag, plus manual setup | Built into every page Snoika manages |
| llms.txt | A contributed llms.txt module | Passes the request through to Drupal |
The modules are free and they're enough if your problem is metadata on pages you already have. Snoika supplies new content for Google and AI assistants when your team can't produce it at a steady pace.
Drupal metatag stays in core
Metatag stays at the center of your Drupal SEO setup after you connect Snoika. The Metatag module is a contributed project separate from Drupal core. On most sites it's the base layer that writes the title and description tags for each node. It also writes the canonical and Open Graph tags. Keep your global defaults and token patterns as they are. After you connect, view the source of a few nodes and confirm the Metatag output hasn't changed.
Snoika serves AI-ready content
Snoika publishes pages that AI assistants can read and quote, and that Google can rank. If you want ChatGPT to cite your Drupal site, give it pages it can fetch and that answer a real question near the top. Allow OAI-SearchBot, OpenAI's search crawler, in robots.txt. Put a direct answer in the opening lines of each page.
Snoika builds researched content maps, so every page targets a question people actually ask. It also tracks how visible your brand is in Google and in AI assistants. Automated workflows draft the pages, and a human review layer checks them. No tool can promise a citation, so watch the visibility data and change the pages that get no mentions.
llms.txt for Drupal sites
You serve llms.txt from Drupal with a contributed module. The file keeps working after you connect Snoika, because Snoika passes that request to Drupal unchanged. llms.txt is a Markdown file at the root of a domain that lists your most important pages for language models. Jeremy Howard of Answer.AI proposed it on 3 September 2024. The llms_txt module from Pronovix adds an admin form for writing the file, and 4,215 sites report using it. Its documentation asks you to set up your web server so a static-file rule doesn't catch the request before Drupal does.
John Mueller, Search Advocate at Google, said on Reddit that "none of the AI services have said they're using LLMs.TXT," and he compared it to the keywords meta tag. The file takes little effort to add and provides no guarantee of citations.
AI SEO without extra modules
You can add an AI blog to Drupal without installing a module because the pages are published at the edge. Snoika creates content for search and AI platforms and publishes it on your own domain. Inside Drupal, the usual route is the AI module connected to a provider such as OpenAI. Then you have API keys to manage and module updates to apply, and your editors still have to build and publish every page. If you need AI drafts to go through Drupal's own editorial workflow, the module route keeps them there. If your editors don't have time for that, publishing at the edge is the lighter option.
Canonical URLs across both systems
Each URL is served by one system, so each URL should have one canonical tag. Metatag sets canonicals on Drupal nodes. When Snoika publishes a new page or takes over an existing URL, check that the live page has a self-referencing canonical. Then check that no Drupal node still competes for the same topic under a different alias.
Sitemap problems arise because your Drupal sitemap module only knows about Drupal nodes. Find out how Snoika's pages get into a sitemap, and confirm in Google Search Console that they're indexed.
No Drupal rebuild required
Your site stays as it is, with its existing hosting and code. You also keep your domain, so every page Snoika publishes counts toward your own SEO. You can plan a Drupal version upgrade on its own timeline, because the Snoika connection doesn't depend on which Drupal release you run. And if you do decide to replace Drupal later, Snoika also builds AI-ready websites from scratch.
Drupal setup steps
Start with the CDN you already use. Cloudflare is the fastest route because you can set it up yourself, and Fastly and the other setups go through Snoika's team. Before you switch anything on, copy the Metatag output from a few key nodes and note their current rankings to establish a baseline. Then you'll have numbers to compare against when Snoika's first pages go live. Snoika handles the new-page research and writing that Drupal SEO modules can't do, then publishes those pages on your domain. To plan the connection for your setup, book a call.
FAQ
Make your pages accessible to crawlers and answer a specific question in the opening lines. Allow OAI-SearchBot in robots.txt so OpenAI's search crawler can fetch them. Citations aren't guaranteed, so monitor visibility data and revise pages that receive no mentions.
Use a contributed llms.txt module and configure your web server so requests reach Drupal rather than a static-file rule. The connection passes those requests through unchanged. The file can list important pages, but it doesn't guarantee AI citations.
Keep your existing drupal seo modules for pages Drupal renders. Metatag still handles metadata, while Pathauto and Redirect handle aliases and redirects. Keep a sitemap module too, because the connection doesn't replace these functions.
Publish the blog pages through a connection at the CDN or web-server level, which serves them on your domain without changing Drupal's code. If drafts need to pass through Drupal's editorial workflow, use the Drupal AI module route instead.
Check their indexing status in Google Search Console and confirm how externally published pages enter a sitemap. Drupal's sitemap module only knows about Drupal nodes, so its sitemap alone won't establish whether pages published through the connection are included.