Feature

posted 31 Oct 2006 in Volume 10 Issue 2

Masterclass: Business taxonomy

Business taxonomy, part I

Taxonomies can be powerful tools. But too often, their complexity defeats the users they are supposed to help. The business taxonomy offers a simpler alternative.

By Zach Wahl

Regardless of how an organisation has designed its taxonomy and regardless of how that taxonomy is supposed to be used, the same issue keeps cropping up: the end users of these taxonomies cannot use them or, more accurately, won’t use them.

In reality, I believe the issue is even more specific in many cases. The real problem with many of the taxonomy designs that have been implemented is that they simply do not make sense to the people who are being asked to use them. A perfect example comes from a US government agency that came to us to ask why the taxonomy it had spent so much time, resources and energy designing, to support its portal and content management systems, was not being used.

The taxonomy had been envisaged, as many are, as a means of categorising all the unstructured content within the agency’s systems and to provide an avenue for users to browse that information. The designers also expected the taxonomy to be used to make search more effective by enabling users to restrict their search results to certain sections of the taxonomy.

Once the system was deployed, however, they found that the users responsible for classifying content within the taxonomy were constantly putting it in the ‘wrong’ place. Moreover, the average end user of the site felt stymied by the taxonomy and inevitably resorted to search as the only means of finding information. When surveyed, end users commented that they could not find the information they wanted and that the system simply ‘didn’t work’.

When the time came to review the taxonomy, the reasons for these problems were immediately clear. The taxonomy had been designed by a crack team of information professionals and expensive consultants, but was only understandable by users with masters degrees in information science.

The taxonomy was detailed – both broad and deep – and included detailed terminology and descriptors throughout. However, the result of these efforts was categorisation for the sake of categorisation. The taxonomy design team had failed to consider the needs and understanding of the average end user and, as a result, had created a barrier to the information those users actually needed.

Taxonomies for the masses

Web technology has made access to information ever-more pervasive – and central to the jobs of more and more people. As a result, more and more ‘average’ end users need direct access to information held in back-end information-management systems.

Typical business users today are also asked to be much more technology and information savvy than in the past – and often with a minimum of formal training.

Just consider: the average user is expected to have an understanding of how information management tools, such as portals, document management and web-content management systems, work. More specifically, these users are expected to know how to search and navigate around those systems to find the information they need.

In many ways, users have risen to these challenges. Today, the average user understands the core concepts behind web information systems and is fairly proficient at both browsing and searching to locate information. However, many organisations are struggling with the basic issue that information management concepts, such as taxonomy design and metadata strategy (historically used only by information science professionals) are now accessible to ordinary business users.

But many organisations have failed to adjust the design and strategy of taxonomies as the audience and uses have changed. As a result, many taxonomies currently found in front-end systems are too complex and far from intuitive for the end users for whom they are intended.

This issue is exacerbated by confusion concerning the uses and applications for taxonomies. This masterclass will address the current disconnects between taxonomy definitions and the organisations that try to build them. It will also introduce the concept of a business taxonomy built specifically to serve the needs of today’s users and applications.

That is to say, a taxonomy that reflects the fact that 99.9 per cent of today’s end users will not be information management professionals or, indeed, care about the classification of their data or where it is kept. They simply want tools that are easy to use and that help them get their jobs done.

Logical concepts

The concept of taxonomies is based on the categorisation of components into a logical structure. Traditionally, taxonomies have been utilised within the scientific and information management worlds in order to classify vast amounts of data into a logical structure.

In that scenario, the users understand the need for such a level of organisation and either know already how it works or are happy to be trained in their use – because their jobs demand it.

Traditional taxonomies are characterised by rigorous categorisation rules, mutually exclusive classification, and exhaustive granularity. As knowledge management tools have become technologically and functionally available to more end users, this concept of the traditional taxonomy has also been applied more widely – and it failed. The average business user – as opposed to a user in the scientific community, for example – regardless of technical savvy or subject matter experience, simply does not receive the intended value from such an offering.

Unlike a traditional taxonomy, designed primarily for the sake of classification, a business taxonomy is designed first and foremost for usability. Whether it will be used to power back-end metadata classification, front-end navigation, or both, a successful business taxonomy must be designed for intuitive browsing by end users.

Design at every stage of the business taxonomy must, therefore, consider whether the average user will be able to understand both the terms and the hierarchy of the taxonomy and react to it in a meaningful and consistent manner. If this is done effectively, the end user will receive a powerful ‘findability’ tool, enabling them to discover information through browsing the taxonomy and view information in an intuitive and consistent manner.

Taxonomies can also benefit search, by acting to limit search results to a specific portion of the taxonomy. These benefits, however, rest on the requirement that the categories within the taxonomy dovetail with what the end user has in mind. These categories need to be easy to understand and act upon.

Core characteristics

With this primary design requirement in mind, the business taxonomy must possess certain core characteristics that differentiate it from a traditional taxonomy.

First, the business taxonomy must be explained with simple terminology that avoids jargon or technical complexity that could confuse potential users. When considering the terms for a business taxonomy, the designers should identify the ‘lowest common denominator’ of user types and build using terms and topics that will immediately resonate with them.

This is especially important at the highest levels of the taxonomy, where all users must be able to immediately understand what they are looking at and where they can go to find the information that they are looking for.

Another characteristic of the business taxonomy is its looser adherence to traditional taxonomy rules. Whereas a strict taxonomy would create a complex hierarchy of mutually exclusive and collectively exhaustive terms, a business taxonomy strives for a flatter, less granular structure that is eminently more intuitive and easier to navigate.

In practice, whereas a traditional taxonomy might have 100 or more top-level nodes that each go 12 levels deep, a business taxonomy might have only eight top-level nodes, none of which will go more than three levels deep. In other words, the business taxonomy sacrifices detail for usability and consistency. Detractors will say that this lack of detail undermines the value of a business taxonomy, but this misses the point: in the business context, usability is key.

Besides, when we consider that we are asking our business users to work within this structure, a simpler taxonomy is clearly the correct choice to ensure the user understands it completely, can navigate it, and can consistently find documents or refer to documents in a repeatable manner.

A flatter structure with less granularity, designed to reduce the number of ‘clicks’ between the users and their desired content minimises the administrative burden of sorting content into the taxonomy. This results in greater usability, and therefore greater findability of content.

Another key characteristic of the business taxonomy is its built-in flexibility, ensuring the existence of a dynamic design to adapt to changing user needs and content. Whereas a traditional taxonomy is rigorously maintained and subject to minimal modification – it has to be due to its size and complexity – a business taxonomy is built on the premise that it must be easy to alter in order to respond to changing business needs and audiences.

This concept supports a methodology of iterative taxonomy design. Organisations must be prepared to support the long-term evolution of their taxonomies so that they can react effectively to both changing requirements, as well as user feedback. Indeed, with the proper feedback mechanisms in place, users will be able to provide taxonomy designers with guidance on how to better organise the taxonomy. In turn, the designers must be prepared to react to this feedback to continuously improve the taxonomy and better serve the end users – they have to be prepared to respond positively, not defensively, to criticism.

Though many organisations fully recognise the value that a business taxonomy offers, they are nevertheless unable to achieve successful business taxonomy designs. There are a number of reasons for this. First, organisations are often unable to break away from the ingrained concepts behind traditional taxonomies.

Even designers fully trained and cognisant of the concepts and values of a business taxonomy tend to fall back to their old habits and default to building ever-more complex structures, forgetting the paramount needs of end users. Another challenge that needs to be overcome in order to achieve a workable business taxonomy is that designers over-analyse their taxonomy efforts, inevitably resulting in project delays and complex taxonomies that are both too wide and too deep.

Approaching take-off

The most effective approach to business taxonomy design is for the organisation to convene a small working group of cross-functional representatives who own and use content throughout their working day. These representatives must be familiar with their own business needs and be able to represent those of their constituents.

After orientation in business taxonomy best practices and goals, the group should be facilitated in such a way as to draw out their stories (or use-cases) for how they think about information and how they classify it themselves.

Starting at the top level of the business taxonomy, this group can define and discuss the primary nodes and then successively dig down into sub-nodes of increasing detail. At every level and at every decision, the group must validate that the decisions they are making serve the needs of the end users.

Though the process may be arduous, the end result will be a business taxonomy that truly reflects the thoughts and needs of the end users. It will, after all, have been articulated by those end users.

Fuelled with intuitive and simple business taxonomies, organisations will foster greater information sharing by breaking down the vertical barriers that currently exist around information repositories.

Individual users will spend less time searching for the information the need and will be able to spend more time benefiting from the information they discover with the help of their taxonomy. With a successful business taxonomy, knowledge management tools will enjoy greater user adoption as a result of increased ease of use. In all, business taxonomies will serve both individual users and their organisations by increasing information findability and system usability.

The forthcoming chapters of this three-part series will focus on helping organisations design and maintain a business taxonomy in order to support end-user needs. The next article will focus on the workshop methodology for designing the business taxonomy and will discuss the formation of the workshop, the exercises to be conducted therein, and the potential risks and challenges associated with the approach. The final part will continue along the design path and discuss guidelines for the long term governance of the business taxonomy to ensure that it continuously evolves to further support end users.

Why taxonomise?

1. Vast amount of unstructured, unorganised, and physically distributed content:

Difficult and time-consuming to find critical information;
Variations in vocabulary and distribution methods obscure important information associations.

2. Enables users to discover and learn:

Quickly and visually presents ‘big picture’;
Creates an explicit and functional map of an organisation's knowledge base;
Enables easy drill-down and browsing of more detailed categories of related information;
Enables users to discover new associations/relationships;
Provides context to information retrieval; disambiguate results; provide more targeted discovery when combined with narrowed search;
Provides consistency in business language and information organisation.

3. Search tools alone are inadequate:

· Are useful when you know what you are looking for;

· Do not provide structure or big picture;

· Do not enable users to make new associations or discover what they don’t know.

Taxonomy development options

1. Wide and shallow

* Requires involvement of all divisions;

* Forces portal engineers to address wide range of document types and file locations early in the process;

* Encourages integrated taxonomy design;

* Potential for ‘so what?’ factor if all information is high level.

2. Narrow and deep

* Enables portal engineers to concentrate on a single issue;

* Single column enables generation of lessons learned before global deployment;

* Ability to get ‘big win’ with one division to serve as champions for portal expansion;

* Only targets one division.

About the author

Zach Wahl leads the knowledge management practice at Project Performance Corporation (PPC). He has led a number of major enterprise-information portal deployments for a variety of organisations, including aero-engine maker Pratt & Whitney, Columbia University, the US Department of Defense, the International Monetary Fund and the US Department of Energy.

Wahl has developed his own taxonomy and metadata design methodologies, authored a series of courses on portal knowledge management and development and is an internationally recognised speaker and trainer on the subjects of ‘e-governance’, portals, and taxonomy design. He can be contacted by e-mailing zwahl@ppc.com.

Founded in 1991, PPC is a management and information-technology consulting firm that focuses on simplifying complex problems for top government and Fortune 500 organisations. PPC’s knowledge management practice has helped more than 120 public and private sector organisations successfully implement the full lifecycle of portals and other knowledge management tools. www.ppc.com.