What you need to provide
To get started with new pages on Similar AI, there are six steps to complete. Most customers finish the full setup within a few days. Here is what we need from you:
- Send us the location of your product feed
- Share a reference product listing page (PLP)
- Set up the proxy
- Set up Google Search Console
- Set up Google Analytics
- Set up internal linking
Your product feed
You need to keep your product information up-to-date to give your users the best experience. Similar AI can accept this information in CSV, XML, or JSON format through an FTP connection, an API connection, or an S3 bucket. It is easier if this information is accessible without a password.
You can put all the basic product details and extra information in one file, if you organize it well. However, keep review information in a separate file because it can get too large.
If you use different files for products, extra details, and reviews, make sure each file has a shared identifier, such as a product ID, so we can match them correctly.
Your information should include everything we need to recreate how your site's search pages look. That means all details about products, images, badges, dates, and so on should be in your files. This helps make sure the new pages look and work just like your own website.
Accepted feed formats
- -Formats: CSV, XML, or JSON
- -Delivery: FTP, API connection, or S3 bucket
- -Reviews: Keep in a separate file from product data
- -Linking key: Include a product ID in every file for matching
Tip: The more information you provide, the better. It means we can create the best pages with the most relevant products.
Share a reference product listing page (PLP)
Please give us a reference URL of a PLP. Point out the elements which are important to you. If there are multiple elements and they do not all appear on a single PLP, feel free to share a few examples to illustrate the point.
In discussion with you, we will agree which functionality to include and which to exclude. For instance, since the goal of these pages is to acquire and convert customers who did not know they would buy from you when they started their search, we often do not include features for logged-in users, such as personal pricing or special promotions.
We will use the reference PLP(s) to replicate your look and feel. We work to ensure that our pages are as similar to yours as possible, but there could be some very small differences.
Set up the proxy
Setting up a reverse proxy means choosing a folder in which Similar AI pages will live, setting up the proxy rules, and restarting your web server. We have detailed guides available for common setups:
- -Nginx reverse proxy integration guide
- -Cloudflare proxy integration guide
- -Magento Fastly integration guide
What proxy setup involves
- Choose a URL folder for Similar AI pages (e.g.
/c/or/shop/) - Configure proxy rules in your web server or CDN
- Restart or reload your web server configuration
Enable Google Search Console (GSC)
For us to get the most data from Search Console, we will set up a BigQuery export for GSC. In order to do this, we will need Delegated Owner access.
If you would prefer not to grant Delegated Owner access, please add analytics@similar.ai to all your Google Search Console properties. We will review and check that you have enough properties.
# Add Similar AI to your GSC properties
Google Search Console > Settings > Users and permissions > Add user
Email: analytics@similar.ai
Permission: Delegated Owner (preferred) or Full
Enable Google Analytics (GA)
We add a cookie to track revenue attribution. Please add the tracking snippet to your category page template. This allows us to measure the revenue impact of the new pages and report on performance.
What to expect
- -We will provide you with a small JavaScript snippet
- -Add this snippet to your category page template
- -The cookie tracks which users arrived via Similar AI pages
- -Revenue attribution data flows into your Analytics dashboard
Set up internal linking
Internal links help search engines discover your new pages and pass authority to them. Choose the integration method that works best for your platform:
Server-side API integration
The most reliable option. Your server fetches links from the Similar AI API and renders them directly in your HTML. See the server-side internal linking guide for full details.
WordPress plugin
If you run WordPress, our plugin handles internal linking automatically with no custom code required.
Client-side JavaScript integration
A lightweight JavaScript snippet that injects internal links on the client side. Quick to set up, but server-side is preferred for SEO.
Frequently asked questions
What does the new page onboarding checklist cover?
The checklist walks you through every technical step needed before Similar AI can start creating and optimizing new pages, including product feed configuration, proxy setup, Google Search Console connection, and Analytics integration. Completing each step ensures the New Pages Agent has accurate data to generate well-structured, indexable pages from day one.
Why does Similar AI require a product feed during onboarding?
The product feed gives the New Pages Agent structured data about your catalog, including attributes, categories, and pricing, so it can build semantically rich pages that match real search intent. Without a properly configured feed, the agent cannot accurately map your products to target keywords or generate correct schema markup.
What role does Google Search Console play in the onboarding process?
Connecting Google Search Console allows Similar AI to monitor indexing status, crawl coverage, and keyword performance for every new page the platform creates. This data feeds back into the Content Agent and Topic Sieve so they can prioritize the highest-impact opportunities in your catalog.
Do I need a proxy configured before Similar AI can start working?
Proxy setup is required so Similar AI can audit your live site and competitor pages accurately without triggering rate limits or blocks during large-scale crawls. Once configured, agents like the Cleanup Agent and Linking Agent can continuously scan your site to maintain content quality and internal link health.
How does internal linking get set up during onboarding?
During onboarding you define your site's URL structure and category hierarchy, which the Linking Agent uses to automatically build contextual internal links across new and existing pages. Getting this right at the start ensures every page the New Pages Agent creates is immediately connected to your broader site architecture for maximum crawlability.
Ready to get started?
Our team will walk you through each step of the onboarding process. Most customers are live within a few days.