This is the third article in the series ‘The Art and Science of Taxonomy Development for Market and Competitive Intelligence’. Click here to view other article
Our brain is trained to think about information in a single dimension. It comes from the mental models that we develop by interacting with the world of physical objects. For example, a physical copy of a job description can be filed in only one cabinet at a time.
These single dimensional models are familiar and simple to understand. Therefore, we organize information in our computer in folders, a single dimension system. For example, we might save information by categories such as prospects, clients, name, type, year, etc.
But this single-dimensional thinking is restrictive. It barely works in managing our limited personal information. And fails miserably when we have to manage large amounts of information.
This is the main challenge in building taxonomies —
our brains are not wired to look at things from multi-dimensional perspectives.
Information is multi-dimensional
Continuing with our example, a digital job description can be organised in across multiple dimensions— by function (HR or Marketing), by business unit (Product A or Service B), by location, or by subjective categories such as senior, mid-Level, or junior positions. It’s difficult for any user to tag all these dimensions manually when working with job postings.
This is very different from our world of physical files and cabinets. It’s tiring to think about information through different mental models, where it can be stored and accessed from multiple dimensions.
Therefore, regardless of education, training, or threats, many users won’t tag all the possible dimensions with the information. Even if some users do, the inconsistent tagging of other users damages the integrity of the whole information system.
Furthermore, the experience is even more frustrating when there are no predefined and logically defined dimensions (taxonomy) to tag the information from.
Organisations have multiple users and use-cases for the same information
In organisations, because there are multiple users in different teams with varying use-cases for the same information, the complexity increases exponentially.
For example, at Contify, we provide market and competitive intelligence data feeds to some of our competitors where we are not directly selling to customers. Our marketing team categorises these companies as competitors, whereas our regional sales team categorises them as customers. (Later we’ll discuss how to resolve such conflicts.)
Building a good taxonomy requires intimate understanding of the users, use-cases and their mapping with the information. It is, therefore, important to acknowledge that it is not as straightforward as it may seem.
Information, users and their use-cases evolve continuously
When new kinds of information enter the system and new use-cases develop in the due course of time, the well-structured taxonomy can degrade into chaos.
It’s a challenge to build a taxonomy that can adapt to not only new information but also adapt to evolving users and use-cases that change with the dynamic market conditions.
For example, one of our customers’ taxonomy had terms to analyse the acquisitions by its competitors. The taxonomy included categories such as geography, employee count and revenue ranges, products, and more. With this taxonomy our customer methodically analysed the acquisitions by competitors over a period of time. And, combining this information with other similarly structured information, such as partnerships and alliances, new office openings, leadership hirings, our customer was able to successfully unfold the strategic plans of its competitors.
Then, one fine day, one of its competitors got acquired by another large company, signalling a potential entry of a formidable new competitor in the ecosystem. This development required a different kind of analysis, and therefore, a different taxonomy structure.
If this customer’s taxonomy wasn’t designed to adapt (discussed later) to new situations, then the whole taxonomy would’ve become less relevant over time. And so would have the applications that depend on it.