Thematic analysis is the most widely used approach to analysing qualitative data. It involves reading through text — interviews, focus groups, open-ended survey responses — and systematically identifying patterns of meaning. Braun & Clarke's six-phase framework (familiarisation, coding, searching for themes, reviewing, defining, and writing up) is the most widely cited guide. Tools like Skimle apply this process to large datasets, surfacing themes across dozens of transcripts in minutes rather than days.
What is thematic analysis?
Thematic analysis is a method for identifying, analysing, and reporting patterns (themes) within qualitative data. The method was formalised by Virginia Braun and Victoria Clarke in their 2006 paper "Using thematic analysis in psychology," which has since become one of the most cited articles in the social sciences, with well over 100,000 citations on Google Scholar.
A theme captures a pattern of shared meaning across your dataset — not just a topic, but a meaningful insight about what that topic means to the people in your study. This distinction matters in practice:
| Not a theme | An actual theme |
|---|---|
| "Price" | "Perceived mismatch between price and delivered value" |
| "Onboarding" | "Implementation complexity as a barrier to early adoption" |
| "Management" | "Loss of autonomy as the primary driver of voluntary turnover" |
Themes are interpretive. They go beyond summarising what people said to capturing why it matters in relation to your research question.
What are the 6 steps of thematic analysis?
Braun & Clarke describe thematic analysis as a six-phase process, which unfolds roughly in this order:
1. Familiarise yourself with the data. Read and re-read your transcripts or documents. Note initial ideas. The goal is deep familiarity before any formal coding begins.
2. Generate initial codes. Systematically work through the entire dataset and label every segment of interest. Codes are descriptive labels: concise, specific, and tied directly to the data.
3. Search for themes. Organise codes into potential themes. Look for patterns across codes — which codes cluster together around a shared meaning? At this stage you are thinking about the bigger picture.
4. Review themes. Check the themes against your coded extracts and the full dataset. Do the themes hold? Do they accurately represent the data? Merge, split, or discard as needed.
5. Define and name themes. Each theme needs a clear definition: what it covers and what it does not. The name should be informative, not just a label.
6. Write up. Produce your analysis as a coherent narrative, selecting illustrative quotes and explaining what each theme contributes to your understanding of the research question.
For a detailed walkthrough of each phase, see our complete guide to thematic analysis.
What are the main types of thematic analysis?
Not all thematic analysis works the same way. The approach you take should reflect your research question and epistemological position.
Inductive thematic analysis lets themes emerge from the data. You begin without preconceptions about what you will find. This approach suits exploratory research, where you genuinely do not know in advance what is important.
Deductive thematic analysis starts with a theoretical framework or a set of predefined categories and applies them to the data. This suits research that is testing or extending existing theory.
Reflexive thematic analysis (Braun & Clarke's later refinement) emphasises the researcher's interpretive role throughout. Themes are not simply found in the data; they are constructed through the researcher's engagement with it. This version is the current standard in much academic qualitative research. See our guide to reflexive thematic analysis for a deeper discussion.
Semantic vs latent analysis. Semantic analysis stays close to the surface meaning of participants' words. Latent analysis goes deeper, interpreting the assumptions and meanings beneath the surface.
Most researchers combine approaches. A study might begin inductively, then use deductive coding to test whether existing theory explains the emerging themes.
When should you use thematic analysis?
Thematic analysis is appropriate for a wide range of qualitative data and research questions. It is particularly well-suited when:
- Your data consists of text: interviews, focus groups, open-ended survey responses, documents
- You want a flexible, accessible method that does not require specialist philosophical training
- Your research question asks about patterns across participants rather than one individual's experience
- You are working with a reasonably large dataset (ten or more sources) where manual reading becomes impractical
It is less suited when you need the fine-grained linguistic analysis of discourse analysis, or when your focus is on a single individual's narrative (where narrative analysis fits better), or when you want to develop a grounded theory from the data upward (where grounded theory methodology is the appropriate choice).
For a direct comparison, see content analysis vs thematic analysis.
What are the strengths and limitations of thematic analysis?
Strengths:
- Flexible and accessible. Thematic analysis does not require adherence to a specific epistemological framework, making it usable across many disciplinary traditions.
- Applicable to large datasets. Unlike interpretive phenomenological analysis (which works with very small samples), thematic analysis scales to dozens of interviews.
- Compatible with both inductive and deductive approaches, so it can serve exploratory and confirmatory research alike.
Limitations:
- Without rigour, it produces superficial results. Simply grouping quotes by topic is not thematic analysis. The interpretive step — constructing themes that genuinely add insight — requires skill and reflection.
- Subjectivity. Different researchers may identify different themes from the same data. This is not inherently a weakness (qualitative research acknowledges interpretation), but it requires transparency about how you coded and why.
- Not appropriate for all data types. Thematic analysis works with text. It is not the right method for observational, audiovisual, or artefact data without prior transcription.
If you are thinking about reliability and inter-rater agreement, see our post on qualitative research sample size, which covers how dataset size affects the robustness of your analysis.
How does AI change thematic analysis?
Recommended reading
What if? Reanalyzing our Qualitative Study with Skimle AI
Thematic analysis on a large dataset — say, 60 interviews of an hour each — produces transcripts running to hundreds of thousands of words. Manual coding of that volume takes weeks. AI-assisted tools can work through the same material in minutes, generating an initial set of codes and themes that researchers then review, refine, and interpret.
Skimle's automatic thematic analysis applies inductive analysis across your entire dataset, grouping segments into themes without a predefined framework. Because every theme links directly back to the source quotes that support it, researchers can trace any finding to the original data and assess whether the AI's interpretation holds. That traceability is what makes AI-assisted thematic analysis trustworthy rather than a black box.
If you already have a framework in mind, Skimle's predefined categories mode applies your structure to new data deductively.
For a practical walkthrough of running thematic analysis with AI assistance, see how to do thematic analysis with AI.
For academic researchers in particular, the key question is transparency: AI assistance is compatible with rigorous methodology as long as you document how you used it, what you reviewed, and how you applied your own interpretive judgement.
Frequently asked questions
What is the difference between a theme and a code in thematic analysis?
A code is a short label applied to a specific segment of text, describing what that segment is about. A theme is a higher-level pattern that groups multiple codes around a shared meaning. Codes are generated during analysis; themes are constructed from the patterns in your codes. Themes require interpretive work — they explain what the codes mean collectively, not just what they describe individually.
How many themes should a thematic analysis produce?
Most published thematic analyses identify between three and seven themes. Fewer than three suggests the analysis has not gone deep enough; more than seven usually means the themes have not been consolidated sufficiently. The right number depends on your dataset and research question, not a fixed rule. If a theme only appears in one or two interviews, it is probably a code rather than a theme.
Can thematic analysis be used for survey data?
Yes. Thematic analysis applies to any text-based data, including open-ended survey responses. The process is identical to interview analysis, though survey responses are typically shorter. This affects the depth of coding possible at the individual response level, but large survey datasets often produce rich themes because the sheer volume of responses compensates.
Is thematic analysis qualitative or quantitative?
Thematic analysis is a qualitative method. It involves interpretation, not measurement. Some researchers count how frequently a theme appears (how many participants mentioned it), but the analysis itself is about meaning, not numbers. Using frequency counts alongside thematic analysis is fine, but the numbers describe pattern, not significance.
What is the difference between thematic analysis and content analysis?
Content analysis counts occurrences of words, phrases, or categories in text. Thematic analysis interprets the meaning of patterns. Content analysis can be fully quantitative; thematic analysis is always interpretive. For a full comparison, see content analysis vs thematic analysis.
Ready to run thematic analysis on your qualitative data? Try Skimle for free and see how AI-assisted analysis surfaces themes across your interviews while keeping every insight traceable to source quotes.
Further reading:
- Reflexive thematic analysis: Braun & Clarke's updated approach
- How to do thematic analysis with AI
- How to write up a thematic analysis
About the authors
Henri Schildt is a Professor of Strategy at Aalto University School of Business and co-founder of Skimle. He has published over a dozen peer-reviewed articles using qualitative methods, including work in Academy of Management Journal, Organisation Science, and Strategic Management Journal. His research focuses on organisational strategy, innovation, and qualitative methodology. Google Scholar profile
Olli Salo is a former Partner at McKinsey & Company where he spent 18 years helping clients understand the markets and themselves, develop winning strategies and improve their operating models. He has done over 1000 client interviews and published over 10 articles on McKinsey.com and beyond. LinkedIn profile
Sources
- Using thematic analysis in psychology — Braun & Clarke, Qualitative Research in Psychology (2006)
- Reflecting on reflexive thematic analysis — Braun & Clarke, Qualitative Research in Sport, Exercise and Health (2019)
- Participant Observation as a Data Collection Method — Kawulich, Forum: Qualitative Social Research (2005)
