Information Architecture Takes Time

When companies organize their information (usually because they are providing a web site for internal or external customers), they often split things up according to existing organizational boundaries. Accounting will have their own area, with accounts receivable, accounts payable, and payroll perhaps occupying different "silos". Sales may have their own sales prospect database -- but it is separate from the customer database that is "officially" shared with the accounts receivable team. Although limiting, this often makes sense because there are already bodies of information organized by functional area, and access control can be assigned based on lists of people that work in those areas.

When conducting an analysis of the information needs of a company, a well-seasoned information architect can start with a notion of how to organize things but have the capability and know-how to make the architecture flexible. The organizational boundaries could be a starting point, but metadata that allows different collections (or buckets) of information to be presented will also need to be defined. Metadata (or information about information) is the backbone of any knowledge management system. The powerful aspect of metadata is that it supports any organization hierarchy (taxonomy) while also letting the system slice and dice the information in a variety of other ways.

You can choose who sees different parts of the same body of information. You can temporarily organize information by project, semi-permanently by brand, or permanently by department. This "permanence" is relative of course. Many knowledge management systems (like Microsoft's Sharepoint) get the information creator to choose a taxonomical category right away, meaning that "no further intervention" is required to organize that piece of information. Still, you can add collections later to categorize the same information differently. What you're actually doing is adding new metadata to the existing data.

On the current Web, the concept of "tagging" has really caught on. This is known as a form of "folksonomy", which means that the consumers of the information apply their own taxonomy and as everyone's metadata is aggregated, an increasingly accurate crowdsourced organization of the information emerges -- yet another way to slice and dice.

At this company, we believe that a comprehensive information strategy needs to embrace three key intertwined elements:

• a pre-defined organizational taxonomy
• a folksonomy (crowdsourced tagging)
• an automated "keyword scanning" taxonomy that derives organization from the content itself (IBM's LanguageWare technology does this)

None of this comes easy; it requires forethought and commitment. Information architecture's return on investment can be measured: productivity gains in accessing the content, litigation reduction when e-discovery is required, and improvement in employee and customer satisfaction when interacting with your company's information. Most companies have found that information architecture is worth the time it takes.