research methods
Thematic analysis
What is thematic analysis?
Thematic analysis is a systematic method for finding patterns of meaning in qualitative data. You read the material closely, label meaningful segments with short descriptive codes, then group related codes into broader themes. It's the rigorous, auditable cousin of affinity mapping: a repeatable path from raw transcripts to a defensible set of insights.
Also known as: thematic analysis, coding, qualitative coding, codes and themes
Prefer to watch?
Watch the recap 0:50
The demo
Five raw interview snippets. Walk them through thematic analysis - code, then cluster, then name - and watch a defensible insight emerge, traceable all the way back to the words.
What this demo shows (text version)
The demo steps through thematic analysis on five interview excerpts. Phase one shows the raw snippets. Phase two attaches a short descriptive code to each ("hidden control", "no feedback", "gave up"). Phase three clusters related codes into candidate themes - for example "couldn't find things" and "didn't trust the system was working". Phase four names the resulting insight, with each theme still traceable down through its codes to the original quotes.
The point is the disciplined, auditable path from data to insight: themes are built from codes, and codes are anchored to real excerpts, so the conclusion can be defended and repeated rather than cherry-picked. It's the more formal cousin of affinity mapping, and like it, the themes are the start of interpretation, not the finished answer.
Thematic analysis turns transcripts into themes through a disciplined pipeline: familiarise yourself with the data, code meaningful chunks with short labels, cluster codes into candidate themes, then review and name them. The discipline - codes tied to actual excerpts, themes built from codes - is what makes the findings traceable and defensible, rather than just the loudest quotes or the team's prior beliefs.
Step through it and a transcript stops being an overwhelming wall of words: each excerpt earns a small code, the codes gather into a couple of themes, and a clear insight falls out the end - with a trail all the way back to what someone actually said. That traceability is what separates analysis from cherry-picking.
The pipeline is the point: familiarise (read and re-read), generate codes (short labels on meaningful segments), search for themes (cluster related codes), review and refine (do the themes hold against the data?), then define and name them. Each theme is built from codes, and each code is anchored to real excerpts - so every conclusion can be traced back to evidence.
That structure is exactly what gives qualitative work its rigour and credibility. Because the path from quote to code to theme is explicit and repeatable, the analysis is auditable, less swayed by the most memorable participant or the analyst's expectations, and easier to do reliably across a team. It's how you answer "how do you know?" about a qualitative finding.
It overlaps with affinity mapping but leans more formal: affinity mapping is the fast, collaborative, sticky-note version; thematic analysis is the more systematic, documented method, often done from transcripts with explicit codes. Both resist imposing categories up front - and both end where the real work begins: interpreting the themes into something you'd actually change.