Effective internal linking: how to work with internal links on the site

How to set up internal linking on the site, correctly distribute link weight, put end-to-end links and design the site structure.

Re-linking a site is linking pages with links. It can be internal, if pages of the same domain refer to each other, or external, if links from different sites are involved.

Site pages have link power - the more links lead to a page, the more important it is and the higher the PageRank. It's not just the number of links, but also the quality of them: Authoritative pages convey more weight.

Simplified on how PR works:

PR transmission scheme, ahrefs.com

Google hid the PageRank values of sites, but did not cancel this indicator - PR remained as part of the ranking algorithm.

This material is about how to properly manage link weight on one site using internal linking.

External linking was discussed in other articles:
11 ways to get backlinks for free
10 experiments with the promotion of satellite sites (PBN)
External links: instructions for use
External links: instructions for use. Part 2

What is internal linking for?

Why do we need interlinking of internal pages:

1. Increases the authority of the page.
   Authority depends on the number of pages that link back to the original.
2. Transfer weight to pages.
   Webmasters use links from authoritative pages to transfer weight to important sales pages that generate traffic.
3. Increases the relevance of pages to the request.
   The robots recognize which request the page to which the link leads is responding to. Of similar pages, all other things being equal, the higher will be the one that is linked to.
4. Speeds up indexing of new pages.
   Search bots first crawl important pages: home, second-level pages, and so on. A link to a new post from an important page will give the bot a signal that it also needs to be crawled.
5. Improves usability.
   Users will be able to read additional information on the topic if they follow the links from the page.
6. Additionally protects content.
   When automatically copying materials from RSS feeds, all internal hyperlinks remain in the text, readers will still be able to go to your site, and it will be easier to find copy-paste .

Ninja Outreach ran an SEO agency London campaign . As a result, organic traffic grew by 40% in a few months. At this time, no other promotion actions were carried out, so the effect cannot be attributed to something else.

Organic traffic graph

The company has sorted the pages into three levels:

• The top ones that rank for the right keywords bring in the most traffic and work as intended.
• The second level - they bring good traffic and a lot of views, are ranked for keywords, but not for all that you need.
• The third level - they do not receive a lot of traffic and do not occupy high positions for keywords, but are well written and useful to readers.

As part of the internal linking campaign, top pages were displayed on all pages, including categories. They linked to second level pages, and those referred to third level pages. As a result, all the pages of the site were linked, the link weight of the authoritative pages was correctly distributed, due to this, the amount of traffic increased.

One team member tested the same campaign on the AvocadoPesto food blog site , resulting in a 20% increase in article traffic.

Internal linking can improve page positions and increase traffic, but it has its own rules.

How to link on the site

Internal site linking is needed not only for the distribution of link weight, but also for usability. The user lingers on the site, follows the links and receives additional information. Relinking consists of several types of automatic and manual work. 

1. Site structure

A properly built site structure will direct the crawlers in the order the webmaster needs and distribute the link weight to the pages that are important for target actions. In order to indicate to the crawler which pages to index, a Sitemap is made - a sitemap with information about the architecture of the resource and the pages needed in the index.

Which site structure to choose

Webmasters can increase the importance of pages by directing link power to them through linking.

Google says the bot is constantly looking for new pages and adding them to the list of crawled pages. He knows some pages, because he has already visited them, and he discovers some by following a link from a known one to a new one. You need to link to new site materials from old authoritative pages so that the bot can find and crawl them faster.

Polish specialist Max Sayrek spent three months experimenting with links: he changed internal links and tracked the crawling of the site by the GoogleBot crawler.

In the structure of many online stores, there are several large product categories with many subcategories. Moreover, in the menu, all subcategories of goods are linked to the main page. In the diagram, the structure looks like this:

Ineffective structure of the online store according to Max Sayrek

This structure is ineffective: GoogleBot sees all links on every menu page, so all pages have the same number of links and the same meaning. In the example in the diagram, the homepage's link weight is evenly distributed across 24 categories with subcategories together, so each page gets about four percent of the homepage's weight.

For interlinking, this site architecture is inconvenient.

Let's say a webmaster wrote an article for his blog and linked it to one of the subcategories. Max's experiment showed that out of several links to the same page, Google takes into account only the first one, ignores the rest. If the link to the subcategory is already in the site menu, as in the diagram, then GoogleBot will ignore the second link from the article.

The structure should be built differently: divide the link weight of the main page into four subcategories, then each of them would receive 25% of the importance and distribute it further.

The correct structure of the online store according to Max Sayrek

Hidden links in menu items that appear only on hover, and hidden tabs with internal links are perceived by the bot as normal, the acceptor can be found by a unique anchor. You cannot hide links from the crawler by placing them in a hidden menu.

Many stores have filters for easy product search. Links that form filter pages are also processed by the crawler. It is recommended to close links with dynamic parameters from scanning so that the bot does not deal with them.

Distribution of links across the SILO structure of the site architecture is considered an effective strategy . SILO implies the distribution of site pages in a hierarchy: content is grouped by topic and structured from general categories to subcategories and individual products.

 

Learn more about SILO content structure and other tips for effective website navigation

Such a distribution of link weight in Runet is called carousel linking - the weight is transferred from top to bottom from the main page to large sections, from there to categories, subcategories and products or articles.

The structure is best illustrated by the diagram:

SILO-structure of content on the site

To combine similar topics, hub pages are suitable - overview pages with a common topic. They provide detailed information about the category to site visitors and group child subcategories, distributing the weight from the main page to them. On the hub, you can post relevant articles and other links to help you learn more about the topic.

Links on one thematic thread can be linked together so that users can go to other pages with products or articles in the topic that interests them.

Linking from authority pages can give weight to important conversion pages.

For example, you can add a “buy with this” block, a block of tag links, links to a manufacturer's category from a product card.

Which links do not convey weight

The experiment showed that the link weight is not transmitted by links via JavaScript - onclick = "window.location.href = 'page4.html'", but crawlers perceive them and can follow them.

The same situation with links via a JavaScript function - class = "js-link" data-url = "page9.html", and the page to which such a link leads, the crawler visited more often than others in this category.

GoogleBot ignored links via JavaScript function with encoded data - class = "js-link" data-url = "cGFnZTEwLmh0bWw =", as well as links with dynamic parameter - page2.html? Parameter = 1.

The link with the anchor - page2.html # testhash also did not work, it did not transfer the weight and the page did not appear in the index at the request of the anchor.

2. Through links

Cross-links - those that are on all pages of the site, as it were, permeate the resource through and through. These include external and internal pages in the header, sidebar or footer of the site.

Website header for roseluxe.ru store

This could be navigation or links to third party sites. They provide the same benefit as external or internal linking, but since the link is on every page of the site, the weight to the acceptor will go from each page.

The positive effect is given by reference to the natural texts , as a website address or company name - Studio "Cloudberry», moroshka.ru.

Cross-links with texts from keys, for example, "buy a bouquet in Moscow", links to other sites in the main navigation of the resource may be perceived by search engines as spam. Such links may fit the definition of "links from low quality directories and bookmarking services" or "the same links at the bottom of pages on different sites." Correcting the text for natural or removing the link will help to remove the negative effect.

Is the link location important?

Google spokesman John Mueller said that it doesn't matter to the search engine where the internal links are located on the page. Their presence and relevance to the user is more important than where they are placed on the page.

On the other hand, Zyppy believes that the links at the bottom of the page get less weight than the top links. It is useful to take this into account if you place links to all sections of the project in the header or footer of the site: information about the company, delivery, payment, catalog sections, access to your personal account and contacts.

Footer of the site of the store labirint.ru

Some experts advise not to overload site navigation with links. The fewer links, the easier it is to track them and direct Google and Yandex crawlers to the most important URLs for crawling.

Wayfair made the site easier to navigate by removing 150 links . There were 51 links in the footer of the site, they removed the ones that were not needed on every page, and left only 20 important for users. As a result, it became easier for users to navigate the sections of the site, they quickly find the desired section and start shopping.

Old and new footer options

Page relevance, user satisfaction, and user satisfaction impacts SEO better than end-to-end links to as many pages on a site as possible. It's better to focus on pages that are important for targeted actions and that answer popular user questions.

3. "Bread crumbs"

"Breadcrumbs" is a site navigation element that shows the path from the home page to the page the user is on. Each caption in an element is a link to the previous category in the site architecture inside the SILO content structure .

"Breadcrumbs" on the lamoda.ru store website

Breadcrumb links transfer weight to the pages and are indexed by search engines. In addition, this is a usability element - it will be easier for the user to return to the general directory or the previous section using quick links from the "crumbs".

4. Block with popular or similar materials

Ring linking method - creating a system of links closed in a ring. Each page links to several previous pages and thus the system closes. Often used for a blog: they display links to articles on a similar topic under the material so that all articles within this topic have links. We recommend using not “Next article” or “Previous article” as the text of the link, but the titles of articles containing keys.

This is how links to materials appear on each page, which transfer link weight to each other. In addition, users may be interested in articles on a related topic and follow the links, from which the behavioral factors of the resource will grow.

Display of other materials on the topic on the article page

Blocks with similar articles are installed using plugins for CMS.

5. Contextual links

Contextual links within the article should first of all complement the material and help the user to better understand the topic.

Design links so that they are visible to users - highlight with color, underline. Do not center links in one place of the material, this is inconvenient for users. Select the most relevant and useful ones so that users click on them and stay on the site, improve behavioral factors. Make links long enough for users to easily access them from their phone.

Link in the text of the article

The page and text of the link must be relevant to the content of the material to which it links. The search bot also accepts links from the image. Use keywords in the title of the image and in the "alt" tag to make it easier for the bot to determine the relevance of the page.

6. Anchors for links

The link anchor is located in HTML between the and tag - Link text

The link text will be the anchor text.

Anchors can be:

• direct, consist of a key phrase, usually high-frequency - "buy a sofa";
• diluted, keywords with clarifications, high and low frequency - “see Dali's paintings of the yellow period”;
• natural - in the form of a domain, a link to a picture, not a keyword.

John Mueller talked about how Google reads the anchor text of internal links to better understand the context.

Replace anchors like “read here” or “about it here” with phrases with keywords so that it is clear what will appear on the link - “ recipe for cream sauce ”, “more about picket fences ”. The bot recognizes link anchors and determines the relevance of the link to the key from the anchor.

For example, you have an article on your website about making pesto at home, and there are several internal links to it with anchors "pesto sauce", "homemade pesto sauce", "make pesto sauce". By these anchors, Google understands that this page is relevant to queries about making sauce and might deserve ranking for these keywords.

This does not mean that you need as many anchor links with a key as possible. This tactic will lead to the fact that the transfer of link weight will be difficult to control, the links will be of poor quality.

As Max Sayrek's experiment showed , if the main word is found only in the anchor, but on the internal one there is no such word at all, a link from the main to the internal one using a unique keyword will give a greater effect to the internal page.