Content analysis vs thematic analysis: which method fits your research?

Content analysis and thematic analysis both involve systematically coding text, and both produce categories or themes. The key difference is what they aim to produce and how the researcher's role is framed. Content analysis is a systematic technique for reducing large amounts of text into categories — it can be quantitative (counting the frequency of content) or qualitative (interpreting what the content means). Thematic analysis is an interpretive method for identifying patterns of shared meaning across a dataset, where the researcher's perspective actively shapes the output. Choosing between them depends on your research question, epistemological stance, and what you plan to do with the findings.

This distinction matters more than it might seem. Applying content analysis procedures to a thematic analysis project (by treating coding as a categorisation exercise, measuring inter-rater reliability, and counting how often themes appear) produces a fundamentally different kind of knowledge. Similarly, applying thematic analysis in a context that requires replicable, frequency-based findings will produce output that lacks the rigour those audiences expect.

What is content analysis?

Content analysis emerged primarily from media and communications research, where scholars wanted a rigorous, systematic method for analysing large volumes of text, images, or broadcast material. Berelson (1952), one of its early codifiers, defined it as "a research technique for the objective, systematic, and quantitative description of manifest content of communications."

The contemporary understanding is broader. Krippendorff's influential textbook (now in its fourth edition) defines content analysis as "a research technique for making replicable and valid inferences from texts (and other meaningful matter) to the contexts of their use." Replicability is central — if two researchers code the same material using the same codebook, they should arrive at substantially the same results.

Hsieh and Shannon (2005) identified three distinct approaches to qualitative content analysis:

Conventional content analysis: Coding categories are derived inductively from the data, without a pre-existing framework. Used when existing theory is limited or would constrain analysis.
Directed content analysis: Begins with a theory or existing research as the basis for initial codes. Deductive in orientation, used to validate or extend an existing framework.
Summative content analysis: Focuses on counting keywords or content and then interpreting their contextual meaning. Bridges quantitative and qualitative approaches.

Even in its qualitative forms, content analysis typically maintains a concern with consistency, codebook reliability, and the ability to report findings in terms of frequency or prevalence.

What is thematic analysis?

Thematic analysis, in Braun and Clarke's influential framework, is an interpretive method for identifying patterns of shared meaning (themes) across a qualitative dataset. In their reflexive version of the method, themes are not extracted from data but constructed by the researcher through active engagement. Coding is an interpretive act rather than a categorisation procedure.

The researcher's subjectivity is a resource, not a source of error to be controlled. Inter-rater reliability is not appropriate or expected, because two researchers interpreting the same data through different theoretical lenses will legitimately produce different analyses — and both can be valid.

Thematic analysis is described as theoretically flexible, meaning it can work within different epistemological positions: realist, constructionist, critical. This flexibility is also a source of confusion — applied without epistemological awareness, it can produce output that neither counts reliably (like content analysis) nor interprets deeply (like reflexive TA is meant to).

What are the 5 key differences?

Dimension	Content analysis	Thematic analysis
Purpose	Systematic categorisation and, often, frequency counting	Identification of patterns of meaning; interpretation
Researcher role	Neutral coder applying a consistent scheme	Active interpreter whose perspective shapes the findings
Treatment of frequency	Central — "how often" is often the answer	Incidental — a theme can be supported by one powerful excerpt
Replicability	Expected; inter-coder reliability (Krippendorff's alpha, Cohen's kappa) is a quality criterion	Not expected; coherence, reflexivity, and depth are quality criteria
Output	Category frequencies, content patterns	Interpretive themes; a rich account of meaning

A sixth dimension worth adding: audience expectations. Content analysis findings are typically presented with frequency data and reliability coefficients — appropriate for policy reports, media studies, and communications research. Thematic analysis findings are presented as rich, argued themes with illustrative quotes — appropriate for health research, psychology, organisational studies, and social science.

When to choose content analysis

Content analysis is the stronger choice when:

Your research question asks "how much" or "how often" as well as "what" — for example, "how frequently do news articles about climate change use uncertainty framing?"
Your dataset is large and you need consistent coding that can be checked and replicated
You are doing cross-study comparison, where other researchers will need to apply the same coding scheme
Your audience (policy makers, regulators, media organisations) expects quantifiable findings
You are analysing structured or semi-structured survey responses and want to report category frequencies alongside your qualitative findings
You need to demonstrate inter-coder reliability for methodological rigour in your field

Open-text survey analysis at scale is a common context where content analysis principles apply well — you are coding hundreds of responses for content, often with a pre-defined framework. See our guide on how to analyse open-text responses at scale for the practical workflow.

When to choose thematic analysis

Thematic analysis is the stronger choice when:

Your research question asks about meaning, experience, or how people understand something
Your sample is relatively small and depth matters more than frequency
You want to produce a rich, interpretive account rather than a category count
You are working within a constructionist or critical epistemological framework
Your audience expects qualitative findings presented as argued themes with supporting quotes
You want the flexibility to explore unexpected patterns that a predefined codebook would miss

For most academic qualitative research, clinical research, and organisational studies involving interviews, thematic analysis or reflexive thematic analysis is the more appropriate framework. Content analysis — even in its qualitative forms — carries an implicit assumption that coding can be consistent across researchers, which conflicts with interpretive epistemology.

Can you combine content analysis and thematic analysis?

Yes, and hybrid approaches appear frequently in mixed-methods research. Two common patterns:

Sequential hybrid: Content analysis is used first to scope a large corpus — identifying the most frequent topics, categories, or framing patterns. Thematic analysis then goes deeper into a subset of the data, exploring the meaning behind the patterns the content analysis surfaced. This is common in large-scale policy analysis or media research.

Parallel mixed-methods: Quantitative content analysis of a large dataset runs alongside qualitative thematic analysis of a smaller subset, with findings triangulated at the interpretation stage.

The key is being transparent about which method you are applying at each stage and why, rather than treating "content analysis" and "thematic analysis" as interchangeable labels for the same process.

A decision framework

If you are choosing between the two, these questions will point you in the right direction:

What is your research question asking? Frequency / prevalence → content analysis. Meaning / experience / process → thematic analysis.
What is your epistemological position? Post-positivist, looking for reliable, replicable findings → content analysis. Constructionist or interpretivist → thematic analysis.
Who is your audience? Policy makers, regulators, communications researchers → content analysis framing is familiar. Academic qualitative researchers, clinicians → thematic analysis framing is expected.
What does your dataset look like? Large corpus of structured or semi-structured text → content analysis scales better. Smaller number of rich interview transcripts → thematic analysis suits depth.
Do you need to demonstrate inter-coder reliability? If yes → content analysis. If reliability is not an expectation in your field → thematic analysis.

If your answer consistently points one way, choose that method and apply it faithfully. If your answers are mixed, a hybrid approach may be appropriate — or the research design itself may need revisiting.

How AI tools handle each approach

AI-assisted analysis tends to resemble automated content analysis: identifying recurring categories, counting patterns, and producing frequency-based summaries. This is useful but does not automatically constitute thematic analysis in the interpretive sense.

Skimle approaches this differently. The automatic thematic analysis generates an initial thematic structure as a starting point for researcher-led interpretation — not as a final output. The statistics view provides frequency and distribution data for researchers who need the content analysis dimension. The two can be used together for hybrid approaches.

For researchers on either side of this distinction, the academic researchers page covers how Skimle fits different methodological frameworks. Market researchers doing large-scale content analysis of customer feedback will find the customer and market researchers page more relevant.

For checking the quality of your qualitative analysis approach, the AI qualitative data analysis checklist covers both content and interpretive approaches.

Frequently asked questions

Is content analysis qualitative or quantitative?

Both. Quantitative content analysis counts frequencies and applies statistical analysis to large text corpora — it is a quantitative method applied to textual data. Qualitative content analysis interprets the meaning of categories rather than just counting them. Hsieh and Shannon's three-approach framework (conventional, directed, summative) covers the range from purely inductive-qualitative to frequency-focused approaches.

Can thematic analysis produce quantitative findings?

Thematic analysis can note frequency (e.g. "most participants described..."), but this is not its primary purpose and should not be presented as a formal quantitative finding. If you need to count and report frequencies reliably, content analysis is the more appropriate framework. Mixing informal frequency claims into a thematic analysis write-up without the reliability procedures of content analysis creates a methodological inconsistency.

What is framework analysis and how does it relate to these two methods?

Framework analysis is a structured qualitative method developed for applied policy and health research. It involves creating an a priori analytical framework, applying it systematically to the data, and producing a matrix that allows systematic comparison across cases and themes. It shares content analysis's concern with systematic application of a codebook, but its output is interpretive rather than frequency-based. It is popular in UK health services research.

Does thematic analysis require a codebook?

Reflexive thematic analysis (Braun and Clarke's approach) does not use a fixed codebook — codes are developed inductively and iteratively as the analysis proceeds. Codebook thematic analysis is a variant that does use a shared codebook, making it more similar to content analysis in its procedures, while remaining interpretive in its epistemological framing. The distinction between reflexive TA and codebook TA is well explained in Braun and Clarke's more recent work. See our guide on reflexive thematic analysis for more.

Can I use content analysis for interview data?

Yes. Directed content analysis is often applied to interview data — for example, using a theoretical framework to code interview responses and test whether the framework applies in a new context. The key consideration is whether the replicability expectation of content analysis fits your research design. If you are doing semi-structured interviews to explore meaning and experience, reflexive thematic analysis is likely to produce richer findings.

Ready to analyse qualitative data with full transparency, whether for interpretive themes or systematic categories? Try Skimle for free — the platform supports both thematic and content analysis approaches, with frequency data and traceable insights.

Related reading:

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