I started reading Information Architecture for the World Wide Web by Peter Morville & Louis Rosenfeld. One of the first books to really address information architecture, Morville & Rosenfeld have recently updated the work to include newer concepts such as tagging. I just finished the chapter on labeling and felt like exploring one of the concepts a little further: card sorting.

While I’ve been acquainted with this concept before, I’ve always found it challenging to generate tangible statistics from the exercise. Likewise, when trying to create an architecture from scratch, it is often difficult obtaining card sort categories from Subject Matter Experts (or perhaps this might just be a weakness of my own). Anyway, an excerpt from the book reads,

"Card sort exercises are one of the best ways to learn how your users would use information. There are two basic varieties of card sorts: open and closed. Open card sorts allow subjects to cluster labels for existing content into their own categories and then label those categories (and clearly, card sorting is useful when designing organization systems as well as labeling systems). Closed card sorts provide subjects with existing categories and ask them to sort content into those categories."

I truly do see the value in a card sort, especially in terms of validating the architecture of an existing site. Where I tend to get stuck is in how to best conduct the sort and quantify the results. Are sorts best conducted in-person or with programs such as those produced by the folks at MindCanvas? While conducting an in-person sort provides opportunities to interact with users (which is always a good thing), software adds value by providing a means of easily quantifying results.

A CMS Wire article reports that Gartner, the research firm, expects the Enterprise Content Management (ECM) global market to increase by 12% per year. This would bring the global content management market to $4.2 billion by 2010. The article went on to state the following:

"The reason for this expected growth can be found on the hundreds of thousands of hard drives inside the computers used by corporate workers across the globe. This “unstructured data” includes word processing documents, spreadsheets, presentations, and rich media files that need control and maintenance."

I always find the large volume of unstructured content organizations produce and disseminate, to both formal and informal channels, to be amazing. I’m very interested in understanding how large volumes of unstructured content can affect the impact of messages deployed by organizations. What do I mean by that? If organizations are producing mounds of unstructured content, this also means that those organizations are producing content in silos. Creating and re-creating content like this can lead to subtle differences over time and at each iteration. This leads to mixed messages for end users: both internal and external to an organization.

When mixed messages are sent out, a “storytelling” effect takes place. One person hears a message one way, another person another way, and so on until the original purpose of the message is lost. I think the only way to resolve such an issue is through some level of training and content management.

I hope I can find some research that gives hard numbers on how unstructured content leads to muddied messages.

The prospect of implementing a structured authoring (or single source, as it’s sometimes called) environment can be quite overwhelming. This is especially true in organizations that have mounds of largely unstructured content. Trying to structure legacy content, in all of its formats, means being able to locate it all and understand the intrinsic relationships.

Because there tends to be a learning curve with stakeholders, implementations should begin with a small group within an organization. If mistakes are made, and they’re bound to be made, the team can quickly adapt in a small but flexible environment. Additionally, the team can create writing guidelines that can later be used throughout the organization. For what is often forgot, or perhaps misunderstood, is that structured authoring is first a methodology, and second a technology.

Structured authoring involves evaluating existing content, breaking it into the smallest possible elements, labeling those elements by content type, configuring the elements into meaningful hierarchies, and linking the hierarchies to related hierarchies. Each element should answer one question. To do this, content types are often developed. These could include such types as the following:

• Topics - Descriptive texts that explain who, what, when, where, and why.
• Procedures - Step-by-step instructions that show how to do something.
• Definitions - Simple or complex lists that define terms.
• Figures - Images that depict physical objects, conceptual processes, etc.

Beginning a structured authoring project in a small group allows the team to mature while on the path to full implementation. For the ultimate success, and future maintenance, of a structured authoring environment depends on the ability of the team to work collaboratively toward common goals.

I read an interesting article the other day via TheContentWrangler, which discussed managerial use of (or lack there of) performance metrics in evaluating writers. The article stated that most technical communications managers have a background in English or journalism and find writing skill and experience to be the most valuable skill set.

Seeing that I have a background in English, I’m curious what metrics can be used to evaluate writing performance. I’ve seen folks in the consulting world use metrics to log “state changes” throughout the day to determine how much “mental firing” they’re putting into their work. As it turns out, most people tend to write, from scratch, for about 15 minutes at a time. There’s also a much quoted statistic that it takes a writer about four hours to write one page of documentation. What other metrics can be used? The amount of documentation produced in a given period of time?

The full article for this topic can be found at < Most Managers Don't Use Metrics As Performance Measure

So I just attended an event at the Pittsburgh Technology Council on Portals, Wikis, Blogs, and other new media technologies. When I arrived, I expected the presenter, Ian Rudy of Plus Consulting, to discuss these topics from a generalist perspective. As the presentation unfolded, however, I soon realized that the event was focused on how each of these new media technologies have been integrated into Microsoft SharePoint 2007.

The presenter touched on a point that interested me: Microsoft integrated a content management system into the 2007 edition of SharePoint. The question on my minds was: what advantage does SharePoint offer over pure content management systems in the traditional sense.

Rudy also touched on how wikis and blogs can be used in enterprises to capture knowledge. A blog, for example, can be used for information that doesn’t quite fit into more formal structures: notes from web conferences, seminars, favorite links someone may come across, etc. Wikis can be used more for dynamic documentation to form a collective knowledge base. IT professionals can use a wiki to update content for new software releases, allowing for minor changes to be quickly integrated.

At any rate, I found the event to be very informative.

I recently read Ann Rockley’s book, Managing Enterprise Content, and had to pause on the section titled “Collaborative Authoring.” In some more traditional authoring environments, writers are asked to take ownership of specific content areas within an organization. Often times this leads to a kind of territorialism among individual authors and creates a siloing in the content creation process. Moving into a unified content strategy with a specific focus on content management and structured authoring only compounds the issue.

Within the technical writing profession thre often exists two very distinct personality types: those that are very integrative and those that are very detail oriented. Rockley writes:

"In a unified content environment, the concept of ownership becomes irrelevant. Instead of the usual cooperative approach to writing that many groups use, a unified content strategy requires true collaboration, which means working together so thoroughly that you no longer own any part, any chapter, any sentence, any procedure. In this way, authors truly become information architects, with more and more emphasis on the analysis and design that drive the content.

What does this mean in the workplace? It means that authors may need new skills, such as information analysis, design, modeling, structured writing, and conflict management to help them in the transition from independent to collaborative work."

I’m curious to know how the more detail oriented technical writers, with years of experience under their belt, adapt to the transition of a structured authoring environment. Many such writers are used to writing in a very linear fashion, a concept that is shattered when writing on an “elemental” level.