As a UX researcher you probably know that tagging is an important approach to get user insights. However, there are many approaches to tag your content.
In this article we will introduce in details the different approaches and explain what value they can bring for your UX research. We will also explain how to create tags for each object category.
Why tag my qualitative research data ?
Tagging, a guaranteed time saver
Tags can save you from having to go through long and tedious mental exercises. With a few tags, you can quickly sort and categorize data for easy retrieval later. These tags also allow you to find your information, no matter where you have stored it.
Tags are the most efficient and flexible way to add data to files.
However, it is important to note that to be productive you must create key tags. Having a mass of tags with poor added value will not help you. The goal is to identify keywords that are consistent with your user’s needs in order to create useful tags to sort your answers efficiently and get insights.
With tagging, enrich your content
Keep in mind that each UX researcher has their own method of semantic analysis. The labeling must be adapted to your objective and your personal method. There are some basic rules to follow but the taxonomy is constantly changing.
Get structured data
Coding/tagging is a process of assigning a specific keyword to various pieces of content. In blogs or articles, its help users find key information and learn more about a certain topic. This is exactly the same for your qualitative data analysis. The only difference is that your tagging needs to be even more specific.
What does “tagging” actually mean? The goal is simply to allow you to draw conclusions from your transcripts in order to interpret and add value to your research.
To do this, you need to tag specific parts of the text. Tagging the most important parts of your data allows you to get structured results. Above all, this process must remain consistent.
For example, if a sentence contains a positive statement, the entire sentence should be tagged with “positive”. Later on, you can deepen your tagging with “subcategories”. Sure, the sentence was positive, but about what ? This funnel effect will allow you to have an organized and structured transcription. To this extent, each statement should be tagged individually.
Provide new information from existing data
Does this seem impossible to you ? Well, you’re wrong !
Tags allow to structure but also to organize information in order to anticipate future questions and/or user problems.
In short, efficient tagging allows you to extract information without having to redo each search session or perform new ones.
But to succeed in extracting relevant information, you need to know how to find the right tag for the right use. We will see that there are two main tags for a good organization :
- Specific tags
Project-specific tags are applied to a specific study and change with each new research. Often they are defined and created as they emerge from the data itself, during the interview. When you are analyzing data for a new research project, you do not know in advance the content of the data you will collect. Project tags help you make sense of the data according to its relevance. You create them based on what the data is telling you, and you quickly adapt them as your understanding grows.
However, project-specific tags are sometimes created in advance. When they are derived from research questions, for example. In both cases, these tags remain applicable and valid only for one project and will not make sense for other studies.
- Global tags
In contrast, global tags provide a common foundation across projects and thus allow researchers to link insights from different studies. A set of global tags is defined in advance with the entire team, as standardized application across projects requires a common understanding of when to apply them.
Global tags mark applicable observations that are generally valid for multiple projects. They are therefore often used to track evidence of already known topics.
Global tags also allow categorizing comments and feedback from users, colleagues, sales or support.
Side note : these two main tags are a valuable foundation for doing your UX research well. They must be combined and will allow you to have a global and complete analysis to perform and obtain quality insights.
How to create and choose the right tags ?
First, to create tags, you have two options. You can create them manually, which requires some work beforehand, or use a specialized tool.
Create tags manually
Setting your tags manually requires some organization. You need to think beforehand about your objectives and the key points you want to target. If you tag manually, you must surely take care of transcribing your exchange as well. All this requires a lot of work and rigor because some essential information could be overlooked.
When you do qualitative user research, there are two ways to tag your notes: deductive or inductive.
Deductive reasoning proceeds from the general to the particular. Scientists use this method to prove theories and hypotheses. For tagging, we can even speak about hypothetical-deductive reasoning. This means putting forward several hypotheses before the start of the UX research. These are defined as a team in order to agree on their relevance. This reasoning then allows each hypothesis to be tested by comparing the expected result with the results of the experiment or observation.
However, this requires a methodology. It is a way to anticipate a project and know what to expect. This tagging will be specific to a particular project and should be meticulously thought out as your conclusions will be based on it.
To begin with, you can establish assumptions along general lines such as :
- Tag for your search
Context (online, offline, social media)
Research methods (analysis, interviews, survey)
- Tags for your organization
Products or services (online store, application, store)
Target group (prospects, persona, consumers, businesses)
- Tags for your user
Stage of the journey (orientation, navigation, purchase)
Features or actions (filters, search, broadcast, edit)
On the contrary, the inductive one proceeds from the particular to the general. This reasoning is done during the interview. The objective here is to succeed in generalizing a series of spontaneous or provoked observations.
In short, it allows to go back :
- from particular cases to a law that governs them,
- from the effects to the causes
- from the consequences to the principle from which they follow,
- from the experience to the theory.
However, you have to be careful with this type of reasoning because a single counter-example can undermine your conclusion.
Exemple : I observe that 100 tigers are brown, so I want to conclude that tigers have a brown coat. However, I learn that white tigers exist, so my conclusion is questioned.
Use a tagging tool
As you can see, tags represent an undeniable added value. But it is a complex process that must be well understood in order to be effective.
This method has been used for a long time, but the manual option is no longer the most effective option. Artificial intelligence has proven itself in this field, and some tools, as Noota, allow to save time as well as to be much more precise in the tagging.
Let’s take a look at automatic tagging !
- Form data extraction
Automatic theme extraction is a huge time saver. It allows you to focus directly on what you need and access the content that adds value to your search.
Moreover, you can analyze one or more interviews at a time. You just have to import your files, and the software will analyze all your data so you can compare them.
And, what about tags?
It’s very easy ! Having already extracted sentences with key elements allows you to create relevant tags. The advantage of these tools is that they save you time but everything remains editable, so you keep control of your project no matter what. You can then add custom topics to access the associated moments.
- The advantage of collection and specific tags
Often the question is how generic or specific to be with tag naming.
For example, if a consumer reports satisfaction with your new product line, how will you name that tag ? It’s best to start collecting all consumer reviews in a generic “new line review” tag that serves as a collector. From time to time, you can browse the tags and see the reviews that come up most frequently. Only then would a specific tag be created for that review.
This approach prevents the tagging from having too many tags. This allows you to have useful tags and to search or filter for frequently mentioned terms and topics.
- Track and evolve tags
When tagging feature requests, it is useful to have a second group of labels for completed requests. When a new feature has been delivered, simply move the corresponding label to the other group.
This way, the feature request group stays up to date and is not cluttered or distorted. Instead of deleting the tag, this allows you to keep all evidence about the implemented feature. You can always track future reactions to the feature and use them for future improvements.
- Benefit from an automatic sentiment score
With clear and efficient graphs, evaluate the positive or negative expressions of your participants during the session on each of the detected topics. Compare them to understand how people feel about each topic.
- View the most frequent keywords
Noota filters the most relevant keywords from your files according to their number of occurrences and lexical field. Click on one of them, to go back to the precise verbatim. You can also transform a keyword into a topic, to extract all related moments.
- Use your question guide
When you run an automated report, you have the option to indicate the questions from your interview. One of our AIs will find them in the text and retrieve the associated answers.