What is information architecture?

This is a combined book review of Information Architecture for the World Wide Web by Louis Rosenfeld and Peter Morville as well Information Architecture: Blueprints for the Web by Christina Wodtke . In a sentence, these two books define and describe information architecture and explain how to put its principles into practice.

Information Architecture for the World Wide Web

Information Architecture for the World Wide Web is a much revised, expanded, and revamped second edition. Its 461 pages are divided into five (5) parts and twenty-one (21) chapters. The first part is devoted to the definition of information architecture and provides an overview of the entire topic. The second part discusses the various components of information architecture in greater detail. Part three outlines processes and methods for practicing information architecture. The fourth part describes how to place information architecture in the working environment (whether it be business or academia). The last part contains a couple of case-studies.

A definition

circles.gif One of the things I liked right off about the book was the "infamous three circles of information architecture." This Venn diagram depicts three interlocking circles representing context, content, and users. These three things make up the ecology of information architecture where users are the intended audience of a website, content is the data/information a website has to communicate, and context is purpose of the website's existence. In order for good information architecture to take place, a concrete understanding of a website's audience, data/information, and purpose are necessary. Just like a architect of buildings, an understand of who is going to live there, what the building is for, and where it will be located is necessary before a plan can be created.

iceberg_s.gif Building on content, context, and users, the authors also illustrate information architecture as an iceberg. Just like an iceberg, the majority of information architecture work is out of sight, "below the water." The work includes the creation of plans, controlled-vocabularies, and blueprints all before any user interfaces are created.

Components: classification, labeling, navigation, and search

Rosenfeld and Morville go on to describe the four principle components of information architecture -- the materials making up cohesive, coherent, and usable websites. They include organizational systems, labeling systems, navigation systems, and searching systems. Each of these systems are described in more detail and best practices put forth. It is interesting to note that each of these systems do not work in isolation of each other. They must work in concert. Each plays in important role in the entire website since each are used to different degrees by users even in one session.


Rosenfeld and Morville seem to compare classification systems to the tables of contents of books. In a table of contents an entire book's content is presented from the widest point of view -- a view from 30,000 feet. It groups (classifies) content into broad categories called parts and chapters. Using a table of contents a person can get an overall idea of what a book is about. Rosenfeld and Morville advocate classifying a website's content in the same way but from the user's point of view, not the hosting institution's. Furthermore, there does not have to be one classification system in place. Things can be organized by chronology, alphabetically, by topic, by format, by task, and by audience all in the same site. The authors also advocate broad and shallow hierarchies over narrow and deep hierarchies. The former requires fewer clicks on the users part and exposes more content than the later.


Labeling systems represent the relationship between users and content. They are the words or short phrases including contextual links, headings, navigation system choices, and index terms. I found this section confusing.


Navigation systems are tools allowing the user to get from one part of a website to another. This is usually implemented with a lot of design in mind. Best practices seem to be tending towards link to the system's home page in the top-left corner of the page identified by the site's logo with the addition of site-wide navigation along the top or down the left-hand side of the page. The bottom of the page then usually includes legal information, a modification date, and a way to contact the author. The remaining real estate is given to local navigation (inside a sub-site) and the content itself. Other best practices include never changing link colors, providing bread-crumbs so people always know where they are in the site, and some sort of site index supplemented with a search function.


Searching is a function near and dear to librarians' hearts. Rosenfeld and Morville describe the issues of search in detail. As they mention, search is not an information technology thing. It is about selecting what to index (and thus what to search), the types of queries to support, how the results are displayed, and what to do with the results. Like a growing number of experts, the authors understand that most people do not and will not input complex search queries. They will enter simple one or two-word phrases and expect the search engine to do the rest. Additionally, the success of search often relies on good, underlying controlled vocabularies, and the book describes their characteristics including faceted classifications.

It is interesting to contrast controlled vocabularies with classification here. While classification is analogous to a book's table of contents, controlled vocabulary terms are analogous to the back of a book index terms. The classification provides a top-down view of content. The controlled vocabulary (index) terms provide a bottom-up view of the content. They expose the details, details easily drawn out by search mechanisms. This why searching a library's catalog is an important way to identify library materials, but browsing the shelf represents an equally valuable alternative. Broad classifications and specific controlled vocabularies are two sides of the same coin.

Processes and methodology

I enjoyed this section the most. It outlined how to put information architecture into practice and advocated a phased approach including research, strategy, design and documentation.


The research phase is used to answer difficult questions about context, content, and users. Questions include but are not limited to:

To answer these questions the authors advocate activities like focus group interviews, log file analysis, content inventories, examinations of mission statements and business plans, environmental scans, surveys, card sorting experiments, and usability tests. Without the research phase, many assumptions about a website will go unsubstantiated or overlooked entirely. The research phase sets the ground work for the balance of website construction. Think "On your mark. Get set. Go!" and not "Fire and then aim."


The strategy phase is used to create a plan. It outlines the results of the research phase and articulates a methods for fulfilling the purpose of the website taking into consideration context, content, and users. By articulating a plan and writing it down the plan becomes concrete and is easily accessible by many people. The plan itself describes what a website will accomplish, who will do the work, how the work will get done, what deliverables will be required, and how long the entire process will take. This is where very highest-level "blueprints" of a website begin to be drawn. These "blueprints" have very little to do with graphic design. They are simply about content and functionality. Another thing to document at this stage, according to the authors, are sets of scenarios on how the website will be used. These scenarios depict real people with real photographs (sometimes called "personas") and help the implementors of websites focused on user-centered design.

Design and documentation

The research has been done. A plan has been articulated. It is now time to put the plan into action by designing and documenting. Design moves from high levels to lower levels. The blueprints (flowcharts) move towards wire frames, and the wire frames move towards graphic designs. All along the way the process the conclusions drawn from the research phase must be kept in mind in order to reduce "feature creep". It is during this part of the process when technology's strengths and weaknesses, advantages and disadvantages are weighed. A content management system may or may not be something to be employed. Existing content needs to be taken into account. Various database tools might be something that is used as a part of the design process. Creating a style guide is also recommended here. This style guide echoes some of the content of the strategy phase, but also explains how the site is organized, why it is organized in the way that it is, and how the site can be extended in the future. The style guide ensures a continued integrity in organization, labeling, navigation, and indexing/searching functionality.

Information architecture: Blueprints for the Web

Wodtke's text is thinner than the Rosenfeld/Morville text, and its a much quicker read. The 348 pages are divided into twelve (12) chapters. The first couple of chapters outline a number of rules of thumb and principles such as:

Like Rosenfeld/Morville, Wodtke discusses content, context, and users, but uses the words users, technology, and business instead. She even puts it another way. Ask who, why, and what about your site.

In the chapter, "Those people", Wodtke advocated user-centered design and outlines the process as:

  1. Figure out whom the site is for.
  2. Talk to those people.
  3. Design the site for them.
  4. Test a prototype of the site with them.
  5. Change the design based on what you learned.
  6. Test the final site with them.

The other chapters of interest go over methods for organizing information, the creation of controlled vocabularies, the art of interface design, and documentation. When it comes to organizing the information, Wodtke differentiates between entertainment sites and informational sites. Informational sites are websites whose purpose is to inform, education, or persuade. Consequently, from the user's perspective the site must address things like:

Similarly, the information architect must:

Again, the answers to these question are implemented through organization, labeling, navigation, and search. Like Rosenfeld/Morville, Wodtke goes into exhaustive detail on how to create thesaurus/taxonomy. Unlike the library profession though, these information architects advocate creating their own vocabulary as opposed to using an existing one -- create a vocabulary that is understandable by the user, not the vocabulary of the hosting institution.

When it comes to interface design, Wodtke strongly advocates the creation of personas -- short biographies of real people who will be using the site. These personas not only list the characteristics of the people using the site, but list the tasks they expect to be able to accomplish.

It is only at this point when interface creation is begun. When creating the interface, it is important to understand:

Wodtke then describes principles of design, global navigation, local navigation with the principles of simplicity, proximity, focus, hierarchy, and using the right tools for the right job kept in mind. Finally, Wodtke advocated writing everything down. This includes site maps, site paths (flowcharts), and wire frames.


These books both discuss the ideas of information architecture. In doing so they have a number of similarities. First, they both strongly advocate an understanding of a website's content, context, and users before any HTML is even thought about being written. To acquire this understanding it is necessary to do research, articulate a plan, and then put the plan into action. When articulating the plan and putting the plan into action, all the authors advocate the use of blueprints and wire frames as preliminaries to any graphic design. Both books mention the "information architecture iceberg" -- a graphic illustrating how the user interface is only part of what information architecture is about. Information architecture is about understanding context, content, and users, and then putting that understanding into practice with methods of organization, labeling, navigation, controlled vocabularies, and search mechanisms. Both books mention ROT ("Redundant, outdated, and trivial") content as an issue with website redesign. The Rosenfeld/Morville book emphasizes methods for making information architecture a part of the organization. Wodtke strongly advocates user-centered design and its principles. The former book is more formal. The later is a quicker and easier read.

Creator: Eric Lease Morgan <eric_morgan@infomotions.com>
Source: This text was never formally published.
Date created: 2003-04-29
Date updated: 2004-12-05
Subject(s): information architecture;
URL: http://infomotions.com/musings/rosenfeld/