Related Searches Recipe Template Focuses on Relevance

The Challenge of Manual and Keyword-based Linking

Sites often rely heavily on manual or keyword-based internal linking, leading to incomplete link coverage, biased link selection, and quickly outdated internal linking structures. Even managing internal links manually for just 50 to 100 pages becomes laborious, time-consuming, and painful. Each time you add or remove pages, you must revisit existing internal links to avoid broken links (404 errors) and maintain relevance. While internal linking greatly improves rankings and user experience, doing it manually is highly inefficient.

The Related Searches recipe solves these issues by using AI to dynamically measure topical relevance, ensuring your links remain relevant, scalable, and unbiased. It works equally well whether your site has 100 pages or 100,000 pages.

Internallinking.io’s Related Searches Recipe template uses AI embeddings to measure topical similarity between the source and destination pages. The AI model generates embeddings that help calculate cosine similarity between pages, resulting in a relevance score from 1 to 10. This score determines which internal links appear, based on your set minimum relevance:

  • High relevance threshold: Displays fewer, highly relevant links but may leave some pages underlinked.
  • Low relevance threshold: Shows more links, but these might be less topically relevant.

In eCommerce, Related Searches link horizontally between category pages. They are a common structure to use to allow users to discover relevant pages, especially when the number of great pages you have exceeds what you can easily fit in your main navigation structure.

For instance, SilkFred has these lovely Related Searches on their Long-Sleeved Dresses page:

SilkFred Related Searches internal linking

This linking module reaches down deeper into the site structure to longer-tailed pages like Long-Sleeved Party Dresses or Maxi Dresses. But it also cleverly adds link to siblings (pages on the same taxonomy level as the source page) and parents. For instance, we can see links to pages like Wrap Dresses or Party Dresses.

Similarly, Wayfair has this Related Searches internal linking block on their Washable Rugs page.

Wayfair Related Searches internal linking

This linking block also reaches out to related pages such as 10' x 14' Rugs or Pink Area Rugs. However, either a lot of pages are getting extra inlinks or Wayfair has turned their relevance threshold all the way down. Although there are links here out to other rug pages, there are a lot of links up to parents and grandparents pages. In addition, because the block has 20 links, some of this block starts to feel a little irrelevant.

When creating a Related Searches Recipe from the template, you define rules specifying:

  • Source pages: Pages you want to build links from
  • Destination pages: Pages you want to build links to
  • Anchor text: Text used in the link
  • Modules or blocks: Where links appear on your page
  • Relevance constraints: Conditions ensuring linked content is relevant, based on AI-generated relevance scores

A major benefit of recipes is that agencies or sites managing multiple clients or markets can use the same Related Searches Recipe consistently. Internallinking.io handles adapting the recipe to each specific site's structure and content, greatly reducing the manual effort involved.

The Related Searches Recipe is our most popular internal linking template, and that's no surprise—people love relevant links. Because users value relevant links highly, so do search engines. Related Searches offers users similar pages that aren’t identical but closely related. Using AI to understand similarity, InternalLinking.io provides options that help users find exactly what they are looking for—even if the initial page isn't a perfect match.

This approach gives users intuitive and relevant choices, helping them discover products or content they wouldn't have thought of themselves, ultimately guiding them to a page that best matches their intent. Related Searches has become the default template most people use to build internal linking for SEO precisely because it aligns user intent closely with available site content.

Related Searches internal linking modules on InternalLinking.io can have as many or as few links are you like. However, typically linking blocks have as few as five links or as many as 15. If you have a lot less than ten, it may be that you aren't using your available linking slots wisely. If you have more than 15, linking blocks can start to look irrelevant or even spammy.

Iterative Refinement and Analysis

Internallinking.io simplifies improving your internal linking step by step:

  1. Crawl the site: First, run a site crawl to collect page data.
  2. Create a recipe: Specify your linking rules using the Related Searches Recipe template.
  3. Build linking graph: Apply the recipe to crawl results to build your internal links.

  4. Review results: Use the versions view, where linking graphs appear as Google Sheets. Sheets show:

  5. Source page

  6. Anchor text
  7. Destination page
  8. Relevance score

  9. Distance (for location-based linking)

  10. Iterate: Adjust your recipe, e.g. tweak relevance thresholds, as needed.

  11. Publish links: When satisfied, [publish links]](integration-on-your-site.md) to the API by clicking "Publish" next to the correct linking version on the Versions view.

Repeat steps 2 to 5 as many times as necessary until your internal linking matches your goals.

Dynamic Updating and Automation

InternalLinking.io automatically keeps internal links up-to-date:

  • Scheduled recrawls: Automatically recrawl your site regularly (monthly, weekly or daily).
  • Automatic publishing: Set your recipes to auto-publish new link updates, ensuring links always stay current without extra effort.
  • Integration options: Results integrate seamlessly with your website through our plugins _e.g. for WordPress, or our standard server-side API. When a visitor loads a page, the latest Related Searches links display automatically.

Automation saves significant time and ensures your internal links are always relevant, without constant manual checks.

Customer Love

“Lyst used Similar.ai to launch a new internal linking module in a week to let search engine users find the pages that they love. In a few weeks, we measured the impact, showing that we had significantly grown the ranking of these pages.” - Simon Dance – Chief Commercial Officer, Lyst

“Similar.ai internal linking increased relevance so that we were able to prove that users were able to find what they were looking for more easily, with statistical significance.”
- Kabeer Badi Singh – Head of SEO, eBay Kleinanzeigen

“We’re working with Similar.ai to leverage machine learning that understands search engine user needs, automatically concentrating internal linking around the pages users love most.”
- Dewi Nawasari – Lead Organic Growth Product Manager, Gumtree.com and Motors.co.uk

  • Enhanced User Experience: Deliver highly relevant internal links that match user intent
  • Improved SEO Performance: Strengthen topical relevance signals for search engines
  • Automated Scalability: Easily manage internal linking for any site size -— whether 100 pages or 100,000 -- removing manual effort

Using the Related Searches Recipe, your internal linking becomes targeted, scalable, and significantly more effective, regardless of your site's size.