Best Practices in Implementing Taxonomy for Enterprise Information Management
The need to classify and categorize corporate information has never been greater. The proliferation of information channels, sources and delivery platforms make managing information a complex business challenge. The issue is further compounded by the fact that, due to the increasing speed of business, information is growing at a rate that far surpasses standard institutional frameworks and controls. Information continues to be recognized as a key source of competitive advantage, and there is an increasing need for the business worker to access relevant information in a timely manner. Information by its very nature is dynamic. Therefore, attempting to set boundaries will result in a cumbersome way to control the flow of information. The key to managing this information is to develop a way to identify, classify and categorize enterprise information. This categorization allows for effective management of content throughout the information lifecycle: capture, storage, retrieval, archival and disposal. The most common way to associate structure around information is to develop and implement a corporate taxonomy.
What is Taxonomy?
Simply put, taxonomy is a way to classify and assign a structure to information. The structure will consist of many levels and sublevels, referred to as nodes and subnodes, each aligned with a specific type or category of information. Not surprisingly, we work with taxonomies every day. A simple example is the folder structures we create to manage computer files. The files are classified on hard drives in a variety of ways to allow us to store and locate information quickly.
Developing a Corporate Information Taxonomy
Developing a corporate taxonomy requires a well thought out approach that addresses not only how information will be categorized, but more importantly, how information units relate to each other and how these will be accessed and retrieved at various points in time.
There are several standard types of taxonomies, the most common being:
- Functionaltaxonomy attempts to represent the business model and organizes information around the services and/or functions the company performs. For example, in a property management company, the taxonomy could be organized by properties, clients and/or leasing activities.
- Organizationaltaxonomy formations mirror departmental functions and thereby are operationally aligned by Marketing, Accounts Payable, Procurement and so forth.
- Topic-basedformations attempt to categorize and label the content by the nature of the content. Examples include financial, policies and procedures, images, contracts, applications and press releases.
- Site map is common in companies that use their enterprise information management (EIM) solution to power their Web sites and in taxonomies that are structured to support Web content.
Taxonomy design, like any other project, requires a well-executed game plan from requirements to implementation. Taxonomy development typically goes through different stages. Figure 1 depicts a taxonomy development lifecycle.
The key is to follow a structured approach to taxonomy development to ensure that in each stage of the lifecycle, there is a conscious recognition of the end goal: to add business value. Taxonomy development is usually an iterative process. Once the general requirements are gathered and the analysis and development of the taxonomy is complete, it is important to perform a pilot of the taxonomy and apply it to actual business scenarios. The lessons learned from the pilot will serve as input to further refine the taxonomy. Taxonomies are entities that generally change as the organization itself changes and matures. Thus, there is no such thing as a final taxonomy, because it can only reflect a point in the organizations life span.
Lessons Learned from Enterprise Taxonomy Development
A systematic approach to taxonomy development will go a long way to ensure that the finished product, the corporate taxonomy, is relevant, useable and provides value to the business. The following are some best practices:
Understand information management for the organization. Prior to developing the taxonomy, it is imperative to understand where the organization has been and where it is going. During the preliminary vision stage, information management may run rampant as its possibilities are pushed to the limit. As momentum is channeled and focused into implementation planning, a sense of reality sets in, inevitably controlling and containing the scope. However, for taxonomy to be truly successful, the entire organizational vision for information management (short-term and long-term) should be considered in the design process. An evaluation of how other IT initiatives impact the scope of taxonomy also needs to be factored. Thoroughly analyze the findings and a vision for the future taxonomy should emerge and serve as a directional force. Employing a holistic approach to taxonomy design will ensure the resulting taxonomy is scalable and able to achieve the long-term vision while enabling information systems to be flexible as business dynamics change.
Evaluate the corporate landscape. Once designed and implemented properly, taxonomies can be a powerful tool in an organization, because they provide a common language for employees to reference and allow employees to leverage content in a unified manner. To ensure the taxonomy acts as a unifying force in the organization, it is critical to weigh the following in the design activities:
- Geographic and localization impacts,
- Organizational structure (centralized versus decentralized),
- Content lifecycles and information flows across the organization and
- Usage model - how it is consumed, frequency of use and types of users (creators versus consumers of content internal and external to the corporation).
The best way to thoroughly analyze these factors and incorporate their impacts is to conduct a series of taxonomy design sessions that involve wide representation from all these areas. Rapid design sessions should have participants from various geographic areas, subject matter experts (SMEs) (who can clearly depict information flows, evolution and departmental handoff) from various functional areas and representative groups of users from the end users to midmanagement to executive levels. The involvement from all these parties makes taxonomy design a time-intensive and reiterative process. However, it yields the most sustainable and well-embraced results and therefore is worth the time, effort and investment.
Establish a systematic taxonomy structure. At the most fundamental level, well-constructed taxonomies depict, organize and classify an organizations content in a systematic and structured manner. As a result, identifying and documenting organizational content through a content inventory is a key activity in taxonomy development. However, content inventories should not be the only input in this process. It is also important to evaluate the impact of organizational initiatives identified in the vision state (even if they are outside the immediate scope) and incorporate these considerations into the design. For example, if the ultimate intention is to leverage the information to provide dynamic content on the corporate Web site, you may want to consider a site map-based taxonomy structure to compliment the portal environment.
Other functionality areas to consider are business process management (BPM) impacts (workflow), reporting and enterprise search, as well as integration into organizational delivery systems such as enterprise resource planning (ERP) or customer relationship management (CRM). Conducting an initial analysis on all these areas will give more visibility into the use and the expectations of the system in advance and will enable the creation of a scalable taxonomy.
Connect taxonomy with metadata. Taxonomy and metadata have a symbiotic relationship. Those who approach taxonomy and metadata development as linear, consecutive steps will have considerable roadblocks with which to contend. Taxonomy structures and metadata definitions are complementary design elements that work cohesively to establish and maintain the information architecture and facilitate access/search capabilities. Together, they form the content model. It is important to acknowledge that this codependency exists upfront. One without the other will inevitably, if not immediately, result in an inadequate design. Taxonomies provide the visual depiction and navigation structure for organizing content that metadata can not intrinsically provide, while metadata provides more descriptive information about the content to provide an alternative way of accessing content across organizational folders. Both work together to yield a synergistic and reinforcing combination.
Clearly determine roles, responsibilities and span of control. With the penetration and popularity of solutions like Microsoft SharePoint in the marketplace, taxonomies have inadvertently taken on a life of their own. While the user-friendliness is applauded, it gives rise to and perpetuates nonenterprise behavior that is contrary to the enterprise concept in EIM.
Many organizations are suffering from content proliferation and duplication and are challenged with controlling, structuring, sharing and mining content. This results in downstream impacts, such as disconnected processes, lack of ownership and operational inefficiencies. Ironically, these are the exact issues that information management solutions intend to resolve. The fault does not lie with the solution providers - after all, technology is simply an enabler of business processes. Instead, it is a lack of strong business practices, specifically as it relates to knowledge management.
In order to control these issues within the organization, clearly define where content can and should reside, set expectations for when and what frequency content should be contributed and updated, and establish the timing for the disposition of content. Also, encourage controlled content lifecycles and embrace it as a corporate competitive tool. Institute formal knowledge management practices and ensure team sites and other functions reinforce and support the knowledge governance model before it is deployed. Most importantly, publicize that a taxonomy is owned by all employees. Proactive behavior within a formal governance structure will ensure the taxonomy continues to provide ongoing benefits.
Develop a taxonomy governance model. One of the most prevalent issues in an information management deployment is that the focus on taxonomy development dissolves once the solution is in production. This is a common mistake, because many implementers view taxonomy design purely as a configuration activity rather than a configuration and maintenance activity. Unlike other systems such as ERP in which configuration is conducted during the implementation and is relatively static once in steady state, taxonomies are dynamic structures. They develop as the business grows, extend as additional technology-related functions are incorporated and morph as the business model changes. Once implemented, it is important to conduct frequent and consistent pulse checks with the business to continually gauge the taxonomys fit and relevance. Armed with this information, the appropriate measures can be taken to adapt the taxonomy to meet evolving requirements. Leverage internal and external SMEs. Taxonomy design can be a very lengthy and tricky process, and as a result, it is more of an art than a science. Taxonomy design is best conducted by someone who has prior experience in the area. The organization should consider leveraging an external SME if needed. In either case, the skills of a specialized taxonomist allows the organization to benefit from their experience and provides valuable foresight into the pros and cons that manifest as the organization matures and grows larger. The conundrum with taxonomies is that there is no right answer. Each taxonomy formation is a trade-off of the pros and cons. The best taxonomy is the one best suited for the operating model and culture. It provides the most benefit while minimizing the impact of the cons.
It is important to continually monitor the taxonomy from a usage perspective to ensure the knowledge management practices are being adhered to. Small clean-up activities may need to be initiated to ensure taxonomy remains clean (removing duplicate objects, reorganizing content, etc.). Over time, usage could diminish as the novelty of information management wears off. Users may need to be re-energized through campaigns or reminders to encourage proper use.
Key Take Away
In developing taxonomy, keep in mind that long-term objectives of an organization will go a long way in ensuring that information is being classified and managed in the most appropriate way. Taxonomy development is not a one time activity but rather a series of ongoing tasks that refine, enhance and improve the baseline structure to meet the changing needs of the business and its ongoing management of information.