Site Navigation & Information Architecture Fundamentals For SEOs

One of the most confusing skills to teach someone who practices organic search engine optimization is called “information architecture” (aka “IA”). I’ve given several talks on the topic in an effort to chip away at some popular misconceptions. Each time the audience has people who refer to themselves as an SEO, IA or usability professional. Typically we’re allotted anywhere from 10 to 20 minutes to discuss “successful information architecture.” Since IA is not the same for SEO and Usability (closely related however), conveying the nuances of how, why, when, where and what information architecture is to each profession feels like a disaster.

Adding to the frustration of communicating what IA is and what it’s for are terms like “mental model,” “taxonomy,” “ontology,” “wayfinder” and many more. If I were to list all the pieces of the information architecture puzzle here, it would take up too much space.

It’s easy to think that IA is all about navigation or even is navigation. It’s not navigation if you think of navigation as offering directions to get from one place to another place. Rather, it’s offering directions in a manner that depends on who or what is asking for them.

For this article, I’m going to focus on information architecture as it is used by someone who optimizes web site pages for search engines. You’ll see the connections to the usability side but the methodology I want to cover now is for the SEO.

Navigation

The screen shot below is of a real web site. What you see is the navigation scheme for the entire website. There are no drop down categories if you click on a sphere.

The internet is rife with websites with sitewide navigation using the words “about,” “services,” “solutions” and “contact.” Sadly, many of them have an SEO assigned to work on them and despite all their keyword and linking work, they ignore the words used to navigate the site. For this site, each sphere links to one page, with no categories. To learn what “solutions” are, one page explains it if you click to read it. (Just love the incentive to go do that!)

The easiest step for an SEO to correct this is to add a keyword to each link label. The most obvious choices would be the brand, specific types of services (consulting, accounting, job finding) and identifying “solutions” (or finding a more definitive word). An SEO skilled in information architecture will take it deeper. This person will want to know what the “mental model” is before picking out any keywords.

Mental model

“Mental models” represent people. And people have brains. Because they have brains, they have information and behaviors stored in their heads. We know humans are creatures of habit and many dislike change. They use commonly known words to describe different things. They’re young or old. Some are deaf. Some wear reading glasses. Maybe they can’t spell. Maybe they only get online when they’re drunk or when they can get to a public library.

Without an idea of who the people are that are expected to use a website, plain vanilla link labels like “about” and “services” are the uninspired standard. But with some mental model data, an SEO can add more zest to their keyword research. I understand that an SEO is often one of the last people to have input on page content during development. This is not a wise practice. For an SEO to create more powerful information architecture for a site, they need to know:

• Who is intended to use the site?
• How they conduct tasks?
• What do they need?
• Variables? Environment, disabilities, age, time, computer experience
• How do they think? Behave?
• User personas
• Card sorting
• User testing during prototype stage
• Analytics
• Market research
• Brain research (neurosciences)
• Human behavior

Why?

Let’s take a moment to consider first-time parents and grandparents searching for baby items. They may want to browse online inside a famous department store. When they land on the homepage, they find confusing navigation. It’s frustrating them because the navigation was laid out without regard to research into the habits and needs of customers. If you were an English speaking person taking a visitor from Japan on a tour of your home, wouldn’t it help to know their language and customs first?

In the example below, there is left sidebar navigation and another drop down menu that appears with the search function. The drop down menu offers redundant or similar link labels as the left side navigation. Which “baby” link do we choose? If you’re looking for a baby gift, do you look under “baby” or “gifts?” For clothes, would you think to try “apparel” or still stick with “baby?” What’s the purpose of the “entire site” drop down menu? Isn’t it adding confusion to the overall user experience?

An SEO might want to add expertise here. Perhaps “clothing” is more commonly used by the site’s customers than “apparel.” Are “baby gifts” something that could be put together rather than in separate categories? What if keyword research strongly suggests those two words are often searched on together?

Mechanics, IA & SEO

Information architecture for an SEO also requires getting into the mechanics of a web site. Consider a site that uses AJAX technology. The URL doesn’t change on AJAX pages. How does this effect IA?

Today, most people understand they can tell the hierarchy (directory structure) of a site by looking at a page’s URL (domain.com/directory/page.html). If that URL never changes, they’re denied this information. And of course, if they bookmark it, they’ll never know what page they bookmarked because it has no accurate identifier.

Text links still rule search engine crawlability. Therefore, an SEO must work out issues with JavaScript and image-based navigation. Link consistency is another area of focus. If a link within content points to a link in global navigation, they should be using the same terms. Navigation, from an SEO perspective, is not simply a matter of creating keyword rich categories or classes for products. What’s happening in the background ties in too, such as sitemap creation, robots exclusions, types of navigation (footer, global, side bar, breadcrumb, etc.), error page handling, internal linking and more.

The key to understanding where the mechanics fit in is to understand and rely on your mental model (who is using the site). An SEO is working on the search engine side, but search engines are dependent on user behavior.

Taxonomies & ontologies

Alan Bleiweiss did a great job working out navigation layouts with categories and sub-categories in Information Architecture – Rocket Science Simplified. He created graphics of example navigation schemes so that readers could visualize relationships, categories and classifications of terms.

“Taxonomy” refers to classification such as categories. “Ontology” means the relationships between classifications or terms. For example, a fruit basket is the classification. Apples, oranges, peaches and mangoes would be related to this classification. If a stakeholder requests that you put “car” into a fruit basket category, it’s unrelated and they need their head examined.

Another example for a taxonomy setup would be:

• Category – Music
• Subcategory – Rock
• Related terms – soft rock, hard rock, etc.

Web sites with enormous inventories face situations where relationships can overlap. Unless the information architecture is worked out beforehand, they end up with orphan pages or hubs (sections with levels) that are completely disconnected from related categories or sub-categories. Before navigation can offer a sense of place, it has to be soundly researched for link relationships, user connections based on user behavior (i.e. baby gifts, baby or gifts), mapped out and coded in the back-end. It’s not difficult to see how important sense of place might be for site visitors, but it’s just as important for search engines. A well designed IA allows search engines to link directly to top level pages that it learns are popular with searchers.

Context, organization and findability

An SEO works with content, whether writing it, creating link labels or linking between pages. While many stop at keyword research, the more skilled will evaluate context issues such whether the site is local or global, what language(s) are used and what terminology choices are the best. An information architect will determine the overall interlinking structure with the goal of creating confidence for both search engines and site visitors by establishing hierarchy and orientation via navigation. Organized page content creation consists of where it goes, how much, order of importance, what goes in and why.

“Findability,” which is another fancy word you may see tossed about, is simply how to make pages easy to search for. Short tail and long tail keyword usage factor into findability. Keep in mind that much of the clues to a fortified site IA are found in mental model research. And best of all, if something isn’t converting as desired it can be examined and changed because user and searcher mental model information is available.

95% of all search engine referrals come from page one of search results pages (SERPS). Organic marketing has been viewed as the “easy” way to market web pages. Some will insist that paying for ads and placement in SERPS is the only way to go. It’s costly and to me, not that challenging. Information architecture is more of a voyage into the art of information communication. Before one syllable of text can be found by either an intended search engine bot or a human person via navigation, we study what they want to know and how they want to find it.

To learn more:
SEO and Searcher Mental Models
Key Information Architecture Concepts Every SEO Should Know