What thematic coding does
An interview can reveal how someone describes an experience, makes a choice, or understands a problem. A collection of interviews offers many such accounts, but the material can be difficult to summarize without flattening important differences. Thematic coding is a way to work through that material systematically. You mark passages that relate to your research question, give them short labels called codes, and then examine how those codes connect to broader patterns.
A code names something you notice in a passage, such as “unclear instructions” or “asking a friend for help.” A theme makes a larger interpretive point, such as “people rely on informal help when official guidance feels hard to use.” Themes are more than frequently mentioned topics: they explain a meaningful pattern in the accounts. The goal is not to make every interview sound the same, but to produce an account that respects both shared experiences and differences.
Prepare the interviews before labeling them
Begin with a focused question. “How do people decide whether to try a new service?” gives you a clearer starting point than “What do people think?” Read each transcript from beginning to end before assigning detailed codes. Note the interview’s context, the questions asked, and any moments where a participant changes or qualifies an earlier answer. A response may mean something different when read alongside the exchange that prompted it.
Keep the original wording available as you work, even if you also create summaries. Remove or protect identifying details where appropriate, and handle recordings and transcripts according to the permissions given by participants. You can code in a spreadsheet, a document, or dedicated software; the important requirement is that you can trace each label back to the passage it describes. A simple record of decisions will help you remember why a code was added, renamed, or dropped.
Create codes that help you compare accounts
Work through a small portion of the material and label passages relevant to the question. Early codes can stay close to a participant’s words: “waited for a recommendation,” “could not compare options,” or “wanted to test it first.” A passage can receive more than one code when it contains several ideas. Do not force every sentence into a category; some material may be background rather than evidence for your question.
As you continue, compare new passages with ones you have already labeled. Ask whether the same code still fits, whether two codes overlap, or whether one label hides an important distinction. For example, “trust” may be too broad if participants distinguish trusting a person from trusting a written explanation. Write brief definitions for codes that you use repeatedly, including what belongs and what does not. Revisit earlier transcripts when your definitions change so that similar passages are treated consistently.
Build themes from relationships, not just repetition
Once you have coded the material, look across interviews rather than reading each one in isolation. Which codes appear together? Which describe a sequence, a contrast, or a condition under which people act differently? Several codes about asking friends, checking reviews, and trying a low-risk option might support a theme about seeking reassurance before committing. The theme should say something useful about the research question, not merely collect related labels under a broad heading.
Test each proposed theme against the underlying passages. Look for accounts that complicate it: perhaps some participants acted without seeking reassurance because they already knew the product. Such cases may narrow a theme, reveal a second pattern, or show that your initial interpretation was too simple. A pattern need not appear in every interview to be worth discussing, but be clear about its scope. Avoid treating a handful of interviews as proof of how an entire population behaves.
Explain how you reached your interpretation
When writing up the findings, state the question, describe how the interviews were coded, and explain why the resulting themes matter. Use carefully chosen excerpts when you have permission to do so, providing enough context for a reader to understand the point. Show variation rather than selecting only passages that make a theme look tidy. Distinguish what participants said from your interpretation of what their accounts suggest.
Finally, reflect on your own role. Your questions, assumptions, and decisions about what to label all influence the analysis. Keeping notes on those choices and considering alternative readings will not remove interpretation, but it makes your reasoning easier to examine. A strong thematic analysis offers a traceable explanation of the interviews: what patterns you found, where those patterns differ, and what remains uncertain.
