Thematic Analysis Software: How It Works & Why You Need It
Most likely, you landed in this blog because you have too much feedback to analyze. You sent out a survey or collected reviews or other form of free-text feedback. Now that you have this feedback in-hand, what do you do with it? How can you identify common themes in responses? How can you create a clear and meaningful report to turn feedback into actions?
Many businesses avoid asking open-ended questions in surveys. Analysis of these comments is very time consuming and expensive. Those that do spend hours sorting through a wall of text in a spreadsheet, coding each text response by hand. This takes precious headcount and a ton of manual effort.
Some end up spending thousands on old-school text analytics software without meaningful outcomes.
Is there a more efficient, less expensive way to derive insight your customer feedback? There is, and it's called thematic analysis software.
In this comprehensive article we cover the following:
If you are only interested in manually analyzing your feedback, check out our guide: How to analyze your feedback in 10 minutes using word spotting. Or, download our toolkit which includes a spreadsheet template to help you get started. Otherwise, keep reading.
What is thematic analysis?
Thematic analysis is a form of qualitative data analysis. The output of the analysis is a list of themes mentioned in text. These themes are discovered by analyzing word and the sentence structures.
For example, let's take these 3 sentences:
There are two key themes here expressed in different words:
- The helpfulness of flight attendants
- Customers needed help setup a baby cot
Thematic analysis can be applied any text. For example, interviews, conversations, product feature requests, open-ended questions in surveys or reviews.
In this article, we'll focus on the thematic analysis of feedback collected at scale. Applying thematic analysis to feedback help quantify themes that impacts business metrics. Knowing this, helps align others on what needs done and gain improvements.
Thematic analysis vs. sentiment analysis
It's not an either-or. In fact, sentiment analysis is often a part of a thematic analysis solution.
Sentiment analysis captures how positive or negative the language is. It finds emotionally charged themes and helps separate them during a review. The example above has one positive and two negative mentions of a theme:
If you only had sentiment analysis, you would know that one person was happy and two unhappy. Thematic analysis tells you what they were happy or unhappy about. Combining thematic and semantic analysis results in better accuracy and nuance.
Earlier, we've shared how thematic analysis compares to sentiment analysis. We also wrote a comprehensive guide on sentiment analysis.
What is thematic analysis software?
Thematic analysis software helps automate thematic analysis. Some software combines human input with algorithmic analysis. More on this below.
How can businesses use thematic analysis software? For example, for finding themes in customer feedback. Do they rate comfort over affordability? Would they pay more for faster service? Thematic analysis software can help you find (and act on) those answers.
The best thematic analysis software is autonomous, meaning:
- You don’t need to set up themes or categories in advance,
- You don’t need to train the algorithm — it learns on its own,
- You can easily capture the “unknown unknowns” to identify themes you may not have spotted on your own.
Want to see an example? You can trial Thematic for free here.
How does thematic analysis software work?
Here’s how thematic analysis software automatically analyzes customer feedback to identify and extract themes.
Natural language processing (NLP)
Natural language processing (NLP) is a subcategory of Linguistics and AI. NLP programs teach computers to analyze large amounts of natural language, aka text.
Thematic analysis software uses NLP to find themes in text. Often, this software also displays that analysis in analytic tools and dashboards.
When a computer attempts to model the meaning of words, sentences, and text, we call it natural language understanding, or NLU.
NLU is a sub-area of natural language processing (NLP). Some NLP tasks, e.g. figuring out a part of speech of a word, might not need to model word meanings for accurate results. But when it comes to thematic analysis, NLU is important. It helps derive the meaning of words used in customer feedback. For example, it can capture that "accommodating" and "helpful" means the same thing.
What about text analytics? This term is a more common way of referring to NLP and NLU in business settings.
How NLP is used in thematic analysis software
The goal of thematic analysis software is to automate theme discovery in text. Natural language understanding (NLU) is an important component in this process. This is different to applying text categorization, which simply puts text into buckets. NLU helps discover themes bottoms up.
This helps us find “unknown unknowns”. These are recurrent points in feedback that you may not have considered. By finding these themes and tracking them over time, you can act on your feedback better.
Algorithms can sometimes have a difficult time parsing negation. For example, imagine a customer responds to your survey with, “There’s nothing I did not like!”. Many solutions will see “did not like” and categorize the feedback as negative.
The best thematic analysis software uses deep learning to recognize positive feedback, even if it’s couched in negative language.
Word embeddings is a deep learning algorithm that finds similar words and phrases. It analyzes occurrences of words across thousands of sentences and spits out a model.
To put it simply, a word embeddings model translates our language (a vocabulary) to a computer’s language (vectors).
Thematic uses a custom word embeddings implementation to turn feedback into a hierarchy of themes:
Why you need thematic analysis software
Now that you know what thematic analysis software does, what about the why?
Here’s how companies can benefit from adding thematic analysis software to their tech stack.
Save time and increase accuracy
When you’re running a business, time is a scarce resource. Thematic analysis software can save your team hundreds of hours a year and prevent them from making wrong decisions.
Many companies still analyze feedback via Excel. This is time-consuming and not scaleable, even for small businesses. Thematic analysis software will help you be more effective.
Thematic analysis software can also help you avoid human errors. When people look at a dataset, we tend to view it through the lens of our own experience and biases. They also might miss something unintentionally.
For example, we once tested Thematic against a human coder, Kate, when analyzing student feedback at a university.
Thematic found that students wanted better food/lunch options. Kate found the same issue, but at a much lower frequency. Why? When Kate looked at the student feedback, she tagged only one key issue per comment. Thematic tagged every issue mentioned in each student comment. Once the university took improved food on campus, student satisfaction increased.
By using thematic analysis software, coders like Kate no longer have to code feedback. Instead, they can use their expertise to interpret the results and drive actions.
Quantify customer feedback
When we talk about quantitative customer feedback, metrics likeNet Promoter Score (NPS) often come to mind. And while NPS scores can be useful snapshots of customer satisfaction, they don’t always tell the whole story.
Why did one of your most loyal customers rate you an 8 instead of a 10? Why did a detractor churn, while a promoter doubled their orders per month? What impact on NPS will we see by taking an action to address a specific customer pain point?
Thematic analysis software helps you find these answers. In turns text feedback into the hard data you need to report and measure the success of an initiative.
“This has made it much easier to get projects across the line, with hard data that we can use to measure success of an initiative." - says one Thematic user on G2 Crowd
“Better yet, we can see how specific themes impact NPS scores!” - shared another
Make data-driven decisions and track results
Thematic analysis software can turn feedback into hard data not only for making decisions but also for tracking progress.
Let's go back to our university example above. A high percentage of students disliked campus food. University put initiatives in place to address this, then they re-surveyed students.
Thematic analysis will show whether students noticed, and what other issues are now on the rise. Here is an example of how Thematic visualizes this in its platform.
3 Examples of thematic analysis software
Depending on your use case, you might want to use a different thematic analysis software. Below, we describe our own Thematic as well as two other highly rated solutions.
We built Thematic specifically for automated feedback analysis. It's best suited for anyone who collects feedback from many different sources such as surveys, live chat, complaints reviews.
And this feedback-focused approach works: 87% of our customers increase their NPS by at least 8 points after using Thematic.
Example of Thematic (video):
How Thematic works: Combining AI with a human touch
We know that every business is different, which is why Thematic lets you combine your unique expertise with powerful AI. All themes are discovered through thematic analysis and are custom for each dataset.
In the Themes Editor, you can adjust themes to make results more relevant to your business’s goals and priorities.
This combination of AI, NLP, and a human touch provides you with a list of themes that is:
- Contextually accurate,
- Varied enough to cover all of the topics in your dataset,
- Meaningful for you and your priorities.
Once you have your themes list, Thematic displays your analysis through customizable dashboards and analytical tools. These tools let you:
- Measure the importance of each theme,
- Compare each theme across different segments of your data, such as demographics or tenure,
- Calculate each theme’s impact is on metrics like satisfaction, loyalty, churn, and spend.
Here’s what our process looks like:
We give you the time and tools to focus on the more exciting parts of analyzing data and reporting on your findings.
DiscoverText is another great example of thematic analysis in action. It's built for academic researchers who need to pull text from public data sources such as Twitter and analyze it quickly. Their data science methods originate in a decade of research with the National Science Foundation.
Like Thematic, DiscoverText understands the value in a human and AI collaboration, emphasising that humans are good at some things, and computers at others. DiscoverText writes that "a consistent back and forth between humans and machines increases the abilities of both to learn."
Here are some of DiscoverText's features:
- Fetches of live feeds,
- Filter by metadata,
- Redact and annotate sensitive information,
- Connect and work with peers in your browser.
Dovetail is a user research platform built for UX researchers who run small one-off research studies. Thematic analysis is one of its key features. It makes it easy to manually analyze text, tag specific parts of feedback with themes and then organize these themes. It's great for collaborating effectively with others and build up reserach repositories.
Key thematic analysis features include:
- Highlighting to tag text,
- Organizing taxonomies,
- Sentiment analysis,
- Graphical reporting.
Conclusion: There you have it!
Now you are a master of thematic analysis software! You understand exactly what thematic analysis is and how it works. You also know how it can help you discover hidden insights in your feedback.
Customer insights and user researchers love the efficiencies thematic analysis software unlocks.
It saves time, money, and is just as accurate as human analysis! (and in some cases, even more accurate).