by Sean McKay and Kade Schemahorn
Teams who do user research, particularly when they’re doing lots of it, are facing a big challenge…
The insights coming out of research are disconnected from each other.
To understand why that’s such a big problem, we need to understand what it takes at a practical level to make good decisions and what barriers stand in our way.
It’s really difficult to get a broader view of what is being learned, to see patterns and make connections across time and across teams.
Once we understand the problem, we’ll look at how you can create a simple taxonomy as a foundation for creating the connections between insights that your organization needs to make good decisions.
What it takes to make good decisions
Good information + experience
When you boil it down, that’s the recipe for good decisions.
- A shared understanding of information that the decision-makers are sufficiently confident in.
- The decision-makers’ experience that allows them to critically apply the information to the present situation.
Experience… that’s a topic for another post. Let’s just assume you’ve all seen a thing or two. Now, how do we get some of that good information?
Well, that’s where “insights” come in. It’s certainly not easy, but we generate insights by doing research.
Here’s the problem…
Each insight captures a particular perspective at a particular point in time. Some insights are more timeless or durable than others, but they all have limitations.
What we’d like to be able to do is combine the perspectives from multiple insights that may have been created at different times or by different teams. Each one helps to widen our perspective. With sufficient coverage, we start to gain confidence that we know what’s going on and can make informed decisions about what to do next.
So, what’s getting in the way?
The barriers that stand in the way of making good decisions
We’ve just looked at how accessing insights and making connections between them is essential to the process of making good decisions.
There are several practical reasons why, in a lot of organizations, it’s difficult to do that.
- Information Overload — Teams often just have a lot of existing research, and they don’t always have the right tools to search through and retrieve information buried in the pile (or piles).
- Buried Insights — Insights often get documented inside PDF reports or PowerPoint presentations. It can be difficult to dig through reports (if you can find them) to find the insights.
- Decentralized, Siloed Storage — Research gets stored in different places and in different formats. This requires people who are consuming research to hunt and peck through multiple systems to get what they need (if they have access).
- Lack of Structured Organization — No matter how well-documented any particular study is, if the larger library of existing research is not organized in a meaningful way, it will be too much work for people to go through it when they have a question to answer.
While there are many remedies to these issues, we’d like to share one that we believe is a foundational step toward greater visibility of and access to insights in your organization—building a meaningful taxonomy to organize research.
How building a taxonomy helps us get to better decisions
A basic definition of taxonomy
When it comes to knowledge management, taxonomies are used to describe what all the documents, images, and other media in an organization are about. Taxonomies contain the subjects that they are related to, organized into meaningful categories.
For our purposes, let’s stick with the basic version of what a taxonomy is. A taxonomy can be thought of as a set of categories and the terms under those categories.
Here’s an example of a very simple, single-level (not nested or hierarchical) taxonomy:
|Simple, abstract taxonomy||Taxonomy using film as an example|
- Term A1
- Term A2
- Term A3
- Term B1
- Term B2
- Term B3
- Term C1
- Term C2
- Term C3
- Establishing shot
Using your taxonomy’s terms as tags to improve access to information
In organizations that have identified the problem of insights being disconnected, one of the first things they often reach for is a tagging system.
“We’ll just tag all our insights! That’ll let us see which ones are related.”
Yes! Tagging can be an effective solution to address the problem of disconnected insights. However, getting good results can be more challenging than you might think.
If you jump right into tagging insights without thoughtfully pre-defining some terms to be used, there are several issues you may run into:
- People are better at recognition than recall. Without pre-defined terms to choose from, you run the risk of people entering multiple terms to refer to the same subject or idea.
- You’ll be at higher risk of typos being entered if there aren’t terms already available to choose from.
- Overzealous taggers may enter tags that aren’t actually relevant to the ways that people might be interested in recalling the insight.
That’s why we advocate for the creation of a taxonomy that has pre-defined terms available for tagging insights. While your taxonomy isn’t likely to be well-refined from the start, even a simple set of terms can go a long way toward helping your team organize insights in a way that is meaningful, specific, and above all useful to your organization.
How to build a simple yet powerful taxonomy
3 ways to develop a taxonomy
- Use existing assets — You likely already have an abundance of resources for sourcing terms for your taxonomy. Existing research documentation, company memos, meeting notes, emails, etc. The vocabulary that your organization uses to describe what they know about users and their behavior is embedded in just about all communication.
- Interview taxonomy users — Product owners, executives, fellow researchers, and other stakeholders. Find out how they are currently looking for information to answer questions they have. How do they find information? What questions do they ask? What terms or phrases do they use? While this approach may take some time, one-on-one interviews can uncover more individualized behaviors/perspectives that may not come out in group settings. It may also be easier to work around individual schedules rather than finding a good time for lots of people to meet at once.
- Collaborative workshops — Involving others directly in generating terms ensures that multiple perspectives are represented, increases buy-in from all involved, and can condense the time frame required to identify terms for your taxonomy.
Doing a first pass at gathering and selecting terms
No matter the approach you take to gathering terms, start by casting a wide net, limiting judgment on what to include.
Once you have a big, unwieldy list of terms, go through it and group like terms. If you are doing a workshop, this is another great part of the process where you can involve others. Take over a room and make this an affinity diagramming exercise.
When all terms have been grouped, go through and identify which groups reflect the subjects that people in your organization might have questions about. In other words, figure out which groups contain terms that might be used to search and browse existing research.
At this point, if you haven’t already, you can create a category name for each group of terms.
The most important thing to keep in mind is ensuring that the categories and related terms you select for your taxonomy are meaningful and specific to your organization. They should reflect the actual vocabulary that is used by people in your organization.
For your first pass, don’t be too worried about finding the “right” categories and terms. Your taxonomy should be living, so build in opportunities to review and refine it over time.
Although each organization will have their own unique taxonomy, here are some example categories that are commonly used:
- Company Goals
- Goal A
- Goal B
- Goal C
- Product A
- Product B
- Product C
- User Tasks
Using emerging terms to refine your taxonomy over time
Taxonomies are often referred to as “controlled vocabularies”. That’s because the terms used in them are pre-defined by whomever is governing or making decisions about the taxonomy. If your system relies exclusively on a “controlled” vocabulary, it can be difficult to evaluate how effective your selected terms are for information retrieval.
In addition to allowing terms to be pre-defined, we suggest using a system that also gives the people doing research the ability to create new terms as needed. It is important to be able to distinguish between the pre-defined terms and the emerging terms that get created as needed.
With the ability to track emerging terms, we can do a first pass at gathering and selecting terms without worrying too much about whether they’re the “right” ones. The best approach to finding effective terms for your taxonomy is an iterative one.
Regularly review the tags that people are using and creating. See if there are more popular terms emerging for any subjects. See if your existing taxonomy needs to be expanded or if there are opportunities to prune unused terms.
When it comes to research, one thing we have in our favor is that many people doing research are already generating new terms as tags while doing data analysis. Handrail automatically distinguishes between pre-defined “official” tags and emergent “freeform” tags. Any new tags that get created while performing data analysis become freeform tags. Any freeform tags can be treated as a candidate to become an official tag.
Are you already using Handrail to manage your tags taxonomy? Let us know how it’s going! We love hearing how people are meeting the challenges of research at scale.