An Information Architect (IA) structures and organizes an online resource like a website. Most of an Information Architect’s work is invisible to most users.

You might notice the effort put into a website by an Information Architect, but is not something you will typically be able pinpoint. Have you ever been on a website where the navigation titles, page titles, and structure of the website made sense, almost as if someone tailored the website for you? Words and terms mean different things to different people, even though a slight overlap in the connotative meaning exists.

How is it possible to categorize words and phrases to where it makes sense to all of these different people? It might just have been the work of an Information Architect!

However, navigation and search do not even come close to the depth of an Information Architect’s work. Information Architects consider three different realms when organizing and structuring an online resource: users, content, and context. They work with users to understand what they want from a website, how you look for information, and your overall experience.

While emphasizing the context, they attempt to incorporate business goals, politics, and other resources into the website. Finally, Information Architects deal with content. Imagine those websites you have visited with hundreds of thousands of pages.

What is the best way to organize and structure the information as well as different types of information, videos, audio, standard web pages, pdfs, etc., so that people can quickly extract the information they want? The answer is simple; hire an Information Architect.


What is Data?

Data is the most abundant element in this trio because data is everywhere. Data is any raw input without some contextual sense to attach to it. For instance, if someone is speaking, each word is data. However, if you included the environment where they were using the word or the person who was speaking the word, you would be slowly casting yourself into the realm of information. Strictly speaking, it is easier to think of data as a number on a blank page. It doesn’t have a solid context, no do we have a caption or an attribute to tell us what this number means. It is just simply a number without any real value attached. Remember in math class where teachers would deduct points when you didn’t include a dollar sign or a unit of measurement. Your teachers might have even put something like “10 what?” to reinforce this point. If you teachers did this, good for them, it means they actually cared about the structure and organization of information. However, data and information often become fuzzy like when you say 2 inches, but don’t express what object is 2 inches long, wide, or tall. No worries, even the experts cannot make a clear distinction between any of these terms.

What is Information?

When this data becomes more organized and structured, it combines into information. Often data and information overlap primarily because information depends on context, but a strong definition for context is very situational. For instance, a picture can be explained as data because it is an initial raw input. But, if you ask an expert, a picture can have an inferred context. It is almost like metadata, but has some inferred contexts, like a background. Information is more easy to think of if you have ever used a database. Information that is contained in a table’s column has an attribute attached to it. This tells us what that value actually means; however, we still do not know what that implies on a larger context without other values. So, if you had a database with a column that was titled
“Visits”, we might try to assume that is visits to a particular web page, but what web page is it? Who was the visit from? How long did the visit last? These questions are typical question you would ask when you wanted to obtain knowledge from the information you are given. Granted, information and knowledge also overlap, but I think it is a slightly clearer distinction that information and data.

What is Knowledge?

Finally, information flows down into the vast realm of knowledge. The point at which information formulates into knowledge is also a fuzzy boundary. Knowledge is the action or communication of the information in a manner that is actually useful to help make decisions. So, if we had a column “Visits” coupled with other columns like “Page Name”, “Username”, and other descriptive categories that we could use to track user visits to a web page. We can run queries on that information to gain useful knowledge. We could see that 100 different visitors viewed the web page about an hour after it was published, which might mean that our users like new content. This might mean it would be useful to publish more new content to increase the number of users to our website. Knowledge can also be argued to be inaction based on information. It is being argued that wisdom is a derivative of knowledge, but it isn’t widely accepted just yet.

Data, Information, and Knowledge relationship


I am going to start out giving you an example of an effective persona that helps you visualize an individual that potentially is in your website audience.

Persona or profile for a typical user of your website's audience

    Key Goals

  • Get relevant information quickly
  • Visualize the field trip
  • Understand the pricing, travel time, and what the operating hours are
    We Must

  • Clearly show location/directions, operating hours, and pricing.
  • Show descriptive yet concise information what education benefit WonderLab can provide
  • Use images to help her visualize the trip

  • Visual learner
  • Is easily distracted by students
  • Easily influenced by pricing, travel time, and operating hours
    We Must Never

  • Overload her with irrelevant information
  • Make it difficult to access information about pricing, travel time, or operating hours

I will not go into much detail as these categories are pretty easy to understand how to create, but the important thing is that most websites have multiple users with entirely different personas. See, once you place all of these personas together you can begin to see patterns emerge to inform you what similarities they have. Those similarities should primarily drive your content and structure of your website.

With our one persona that we created, we know that she likes fast access to information, which means we probably shouldn’t create a website with a lot of depth. We might create a broad navigational system where she will not have to click more than 2 or 3 links from any page to reach another page. We know that she likes to see important information like location, directions, operating hours, and pricing. We might consider placing that in a sidebar to clearly portray that information on every page, which means she will find that information on every page. Finally, we notice that she loves images, which means we should provide a lot of graphics to complement the text. Of course, personas might conflict, but you should attempt to produce the best possible fit for your website to all of the different personas. This is precisely where the Information Architecture field turns into more of an art than a science. Use your best judgement, and always remember your users are what is the most important to the success of your website.


A representation itself is an abstraction from an object. With that said, the state of a representation is how far it is deviated from the initial object that it represents. In terms of usability, we often study the effects of a metaphor to portray an idea, and I would state that metaphors are very similar to representations. I actually believe them to be the same thing. People tend to thing of representations as images or graphic constructions that simplify a previous object by removing the less relevant aspects of the initial object. Despite the tangent, representations are very effective ways that you can tailor your content to your visual learners. Also, representations can reduce the complexity of a situation by removing the less important aspects.

Think for a minute about how many representations that guide you through life every single day. If we didn’t have representations you would be in complete information overload. For example, in an average week, you might use Google maps or a GPS, which shows you the roads to get to a certain location. That is awesome, but think of how many aspects Google leaves out in delivering you the representation. While they do have the option to give you a satellite view where you can see the trees, mountains, etc., but they don’t include it initially because it would just confuse the user. Let’s take a look at a critique of a representation found on a refrigerator:

A critique of a Graphical Representation

The graphic, admittedly, is a very sophisticated system of representations of how this particular refrigerator operates. Without previous abstracted representations, one still may find difficulty in interpreting the knowledge embedded inside this graphic. However, it is probably expected that the user already possesses knowledge of the internal workings of a refrigerator. From the representation, I have developed a relationship between the programming concepts of inheritance and objects to relate to the understanding of representation and abstraction (Abstraction is also a term used in OOP as a slightly altered relationship to another object). This graphic can be deduced to a complicated meta(meta)representation.

One of the obvious and simple representations is the magnet inside the compressor utilizing the characters S and N. The more complex (meta) representation incorporates the basic magnet into the entire compressor. While it is interesting to analyze the representations and abstractions, they all collaborate to give us this greater understanding of the internal process of a refrigerator. Granted, the graphic lacks many fundamentals of a great representation, but it does immediately register as a process. With some external knowledge, we can guess that the representation begins shortly before the thermostat, and ends slightly after the compressor.

A final important interpretation from this exercise is not to forget the human constant when designing representations. It isn’t abundantly clear what knowledge can be inferred from this graphic. Ultimately, the graphic demonstrates how a representation system is still dependent on the people and context.

What have we learned from the refrigerator representation? We should have noticed that the representation omitted a considerable portion of context. You don’t see an actual refrigerator do you? When the person who created this representation was designing this graphical representation, they clearly believed that the image was useful for those individuals who might need to gain information about how this refrigerator worked. I would assume that the average person is clueless about what this graphic is intended to explain, which tells us that all graphical representations are dependent upon the user’s experience. Sure, an electrical engineer probably could spend the day explaining this to us, but without him, the graphic is just an image without meaning. When designing your representations, think of the knowledge dependencies that are required to understand it. Blindly creating representations is not exactly a useful practice and might actually be counterproductive for your user. Think to simplify from your users’ viewpoints.

Organizational Structure

Organizing and structuring are basic human fundamentals. Let’s face it, we are pretty darn good are categorizing and ordering things that have no apparent order. However, the internet has become a giant and we are swimming in it’s content. Humans don’t always categorize the same way as others do, which is why it is so important to understand the basic categorizing and organizing principles.

There are many different ways you can categorize something; alphabetically, chronological, by function, etc. The important thing about how you categorize is how people will infer your categorization, which means you need to make it easy for them to understand and predict. Don’t make your users think! Let’s take a look at 2 different shoe retail websites to break down how they categorize their navigational structure:

The first website I have chosen is www.payless .com, which is a company that sells various types of shoes. Using the LATCH principles, I believe that they have utilized “category” as their organizational schema. The primary navigation is categorized in such a way that it is grouped by individuals who would purchase the shoes. However, I must admit I am not entirely certain of the ordering of the elements. Perhaps, it is directly correlated with the most commonly visited sections. While it is a generalization, I would venture to guess women would be the most frequent visitor of the website. Of course, this order based on frequency would fade as we move the right. The secondary navigation is ordered by these vague categorizations of “Trend” and “Style”, but the subcategories of those shows a much more clear order using an alphabetical ordering.

The next website is www.shoecarnival.com, which is also a company that retails various types of shoes. Interestingly, this website has a very similar primary navigation structure as the previous website. Again, I assume it is based on frequency, as it does not seem that many other viable options exist. However, this website does have a significantly different secondary menu. Most of these items are found in the Payless website’s secondary navigation, but Shoe Carnival has reduced the number of total categories. While seeing that this categorization does not follow an obvious organization, I would also venture to assume it is also grounded on frequency.

As a final note, I believe that Shoe Carnival highlights a great fundamental concept that Payless does not have. Shoe Carnival provides additional browsing parameters based on categorizations like size, width, and color. They execute the options perfectly as it is not overwhelming to the user because all of the categories have less than 8 items (with the exception of size), unlike Payless where I had to scan through 14 different options under one section.

Do you see the actual difference in the quality of the organization? Hopefully, the websites do not change dramatically since I wrote this review, but it should help you understand how Shoe Carnival does a better job organizing the content. Think of ways that you can draft similar categorizations by making them more simple and intuitive for your user. Users do not particularly enjoy clicking a bunch of links in a blind fashion to find the content they want. They would prefer to feel that they are clicking the right link to get to the content they want on their first attempt.