Site architecture is one of the fundamental SEO (search engine optimization) aspects that need to be considered at the first stages of web development. Setting in place a great site structure -that is accessible both to users and search engines- allows any business to develop a SEO-friendly attitude since the beginning. From the point of view of users, it means that a site easy to navigate will be more easily visited, referred and shared. From the point of view of the search engine, a clear structure means that the site is crawled more easily and therefore indexed more rapidly. So…how do you build a great site architecture both for users and search engine?

Determine your website theme and create a Silo

Silo Site Organization

As Google’s mission is “to organize the world’s information and make it universally accessible and useful.” a high ranking site must provide:

  1. Useful information that goes beyond the site design and layout
  2. Clear structure and language (both for users and search engines)

According to Bruce Clay, a site can be compared to a jar of marbles of different colors. If the jar contains randomly mixed marbles it is difficult to determine how many of each the jar contains. In a similar manner, a site that does not organize its content in a rationale manner, will cause users and search engines to struggle to navigate and identify the site’s content. However, if the site is divided in themed categories in a similar manner in which a jar is divided into separated sections of colored marbles, users and search engines will be able to select a specific theme while ignoring the rest. This is what a Silo is. But…how do you build one that ranks better among equal sites?

  1. Know what your site is about
    • Find 10-100 keywords that best describe your site
    • Search on Google for those keywords. Do you come up with a site that is similar to yours?
  2. Create a clear theme
    • The major theme will be your homepage at the top of your structure
    • A series of categories will branch out from your homepage as links/URLs
    • A series of related sub-categories will branch out from each categories
    • Single item pages will branch out from each sub-category
  3. Create a Silo
    1. Directory Silos
    2. Virtual Silos
  • Directory Silos
    • Content is organized in separated categorizes with no cross-linking
    • Each category has a distinct descriptive heading and URL
    • One silo may contain multiple sub-silos (don’t go deeper than 3-4 levels)
    • Each silo should contain general keywords (shoes) and long tail keywords (high heels sandals)
  • Virtual Silos
    • Multiple silos are connected by hyperlink
    • Supporting pages are linked to the category they belong to plus to other supporting pages for that theme.
    • This type of cross linking establishes and reinforces the theme.

Build horizontally not vertically!

Deep and Flat Site Architecture
Deep (left) and Flat (right) site architecture

Most SEO specialists argue that site should be build with a flat structure in mind. That means that from the top level (Homepage) to the bottom one you are always few clicks away. This type of hierarchy implies that there are many top level categories and less bottom level subcategories. In the opposite case of a deep or nested site architecture, you have multiple levels with few categories and subcategories at each level. In the first instance your site structure would look like a short pyramid with a large base, while in the second instance, the structure will be much taller with a narrow base.

Why does an horizontal, flat site architecture help SEO?

In terms of user experience, visitors will more easily reach any content if they are few clicks away from any page. Visitors do not like to waste time in finding the information they are looking for: too many clicks might discourage a user to keep digging into your site. In the long terms that means losing visitors and therefore make a website useless (unless you use it as your personal diary!). In terms of search engine, a deep nested site structure is more difficult to be accessed by engine bots as once they have reached your site, they need to follow a long chain of links. In addition, the higher a page is in the hierarchy of a site, the higher the priority which could eventually help ranking.

In a flat site architecture there are more links per level…is that bad?

As there are fewer levels, a flatten structure will end up having more links per level compared to a deep nested site architecture. According to Google best practice, too many internal links per page is not considered SEO-friendly and some could end up being ignored by the search engine crawlers.

The Pagination issue of a flat site architecture

In a flat site, there are fewer subcategories and therefore in the case of a  e-commerce website, more products and therefore more pages will end up in the same group. If you decide to limit the number of products per category to limit the number of links coming out from a single category page, you could decide to opt for a paginated series option. The pagination can come in three form:

  1. Pagination with rel=“next” and rel=“prev”
  2. Pagination with “View All”
  3. Pagination with “Infinite scroll”
  • Pagination with rel=“next” and rel=“prev”

In this case a page shows a limited number of items and to view more a user needs to click NEXT or (to go back) PREV. What are you saying to the search engine by adding HTML link elements rel=”next” and rel=”prev” ? First, that the search engine should consolidate indexing properties, from the component pages/URLs to the series as a whole; second, that the users should be sent to the first one of the series as that stands out as the most relevant. You can decide to opt for such a pagination when a single article or forum thread cannot be displayed in one single page. In some cases galleries and lists can be displayed in a manner that every item has its own page. In some instance you might decide to add a nofollow to all the pages beyond page 1 in case you consider that page one is the most important and the rest could be ignored by crawlers. In my option, this could be risky as these pages will no be indexed. In this case, go to your Google Webmaster tools page, open the Parameter Handling section and set the paginated page parameter to “Paginates” and for Google to crawl “No URLs” (otherwise just leave the setting to “Let Googlebot Decide”) or you can tell Google Webmaster Parameter Handling to set the paginated page parameter to “Paginates” and for Google to crawl every URL.

  • Pagination with “View-all”

This is the option that search engines prefer. Therefore, if your site offers either the pagination with rel=“next” and rel=“prev” and the “View-all” option, Google will prefer the second option. However, the problem with this option is the high latency that a page would display in case it is required to load many items. In that instance, webmasters should opt for a multiple page pagination to improve the loading time and the overall user experience. If you want to make explicit that the View-all is your favorite option, you can add <link rel=canonical href=”view-all.html”> to each of the multiple displayed pages. Conversely, if you prefer the search engine to ignore View-all, just add noindex. This can apply when users prefer the multiple page display. In this instance the search engine will consolidate the indexing properties between the components and send users to the most relevant page (page 1). Finally, set the paginated page parameter to “Paginates” in the Parameter Handling section of Google Webmaster Tool to tell Google to crawl every URL. As the View-All page might become too heavy to load you could opt for the “Pagination with Infinite scroll down” solution.

  • Pagination with Infinite scroll down

This type of pagination allows users to scroll down a massive amount of content with no end in sight. This time of solutions can sometimes make users feel like lacking control on the data they are exploring and less likely to find information that they find useful as overwhelmed by the amount of data that keep loading. This solution can become even more frustrating when users revisit the same information a second time around. By being unable to pin point the location of the item they were interested in, users are required to skim through the same list all over again like the first time visit.

Pagination vs Infinite Scroll, who wins?

The main issue with Infinite scroll is the lack of control the user have over the amount of content loaded and its location along the page. Conversely, pagination can create limitations on the crawlability of multiple pages, limit the page juice transferring to pages they are linked to or dilute the page juice among paginated pages. Finally, coding errors could make paginated pages look like duplicated content to search engines or like low quality pages due to thin content.

A new solution: the Faceted Navigation Option

A very user friendly navigation menu allows users to filter items by multiple categories such as price range, size, color etc. Personally, as a serial shopper, I love this option as it allows me to go through shoes of my size only or bags of my favorite colors. However, it could also not be Search Engine friendly as it appears to create duplicated content and therefore making the crawlers take more time in indexing those pages. One solution is to make all unnecessary URLs links rel=“nofollow” and consolidate indexing signals  by adding rel=”canonical” from the unnecessary URL to a superset URL. You can even decide to disallow unwanted URLs in robot.txt and therefore let all search engines freely crawl only good content. Finally, you can even decide to send the unnecessary URLs to a separate host in order to create a high and low ranking site.

Navigation that reflects user behavior

A good navigation menu creates a glue that sticks together all the top level pages. Today every user expects to find a universal menu located either at top of the page or on the side from which they go back to top level pages from any page they are visiting. These pages will also be identified by search engine as the most relevant pages of your site and therefore will receive the highest ranking. But how do you decide which pages are the most important?

  • Keywords analysis
    • If the users come to your site because of specific search terms, optimize those pages to make easy for them to find immediately that type of content
  • Most visited pages
    • If specific page get the most traffic, add links to those pages on your homepage (the same can be applied to exit pages)
  • Optimize your categories by linking pages together
    • Add a vertical linking with breadcrumbs
    • Add horizonal linking to recommend similar products
  • Be wary of Drop Down menus
    • They must be in HTML to allow search engine to crawl them
    • They must be excluded in mobile apps as they do not work on such devices
  • Add Filters
    • Create filters to allow users to view only products/items they are interested in
    • In an ecommerce site, size should be listed before price
    • They help user experience and are crawled by search engines


Please leave me a comment and let me know how I could improve the article.

Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.