This article dives into how to run thematic analysis using NVivo, powerful qualitative data analysis software. You'll find a quick refresher on the six phases below, followed by where NVivo tools—memos, code hierarchies, coding stripes, Framework Matrices—and the AI Assistant add-on fit into each one, so you can move from raw transcripts to well-documented themes without cutting corners on rigor.
The six phases of thematic analysis
Thematic analysis is a method for identifying, analyzing, and interpreting meaningful patterns — or themes — within qualitative data. Most commonly associated with Braun and Clarke, it follows a six-phase process that can be applied inductively (where themes emerge from the data) or deductively (where analysis is guided by existing theory), and works with interview transcripts, focus group discussions, open-ended survey responses, and written documents. Want the full picture, including how thematic analysis compares to discourse analysis, content analysis, grounded theory, and framework analysis? Read our complete guide to thematic analysis.
The six phases are:
- Familiarization — Read and re-read your data (transcripts, audio, video) to get a feel for its overall tone, flow, and the range of meanings present, before you start formal coding.
- Coding — Label meaningful segments of the data, capturing content participants stated explicitly (semantic codes) or underlying meanings you're interpreting (latent codes).
- Generating initial themes — Group related codes together to spot patterns, organizing them into broader categories that reflect shared meaning across the dataset.
- Developing and reviewing themes — Check that each theme is consistent, distinct, and grounded in the data; some themes may need to be split, refined, or dropped.
- Refining, defining, and naming themes — Sharpen each theme's definition and scope, and give it a clear, concise name that sets it apart from the others.
- Writing the report — Present your findings, supporting each theme with representative excerpts and linking back to your research question.
These phases can be done manually, but qualitative data analysis software (QDAS) like NVivo makes each one faster and easier to document.
Where NVivo fits into the workflow
NVivo won't do the thinking for you, but it gives you structured tools for organizing, coding, and documenting your process at every phase:
- Familiarization: Import your sources and add memos to capture initial impressions, flag standout passages, and annotate the text directly—all before you apply a single formal code.
- Coding: Highlight text and apply new or existing codes, organize them into parent-child hierarchies for complex or layered themes, and turn on coding stripes to visualize overlapping or related codes across multiple passes.
- Generating initial themes: Group related codes using sets or folders, run matrix coding queries to see how different codes interact or overlap, or build a Framework Matrix—a grid with cases as rows and themes as columns—to bring all the coded data for a case and theme together in one place.
- Developing and reviewing themes: Retrieve all data coded under a theme and review the segments side by side to check whether it holds together or is too broad or vague, and use memos to document your reasoning as themes evolve.
- Refining and naming themes: Generate summary reports with theme names, definitions, and representative quotes. If you built a Framework Matrix earlier, the AI Assistant can draft a summary directly within each cell—a starting synthesis of a theme across cases that you review and refine.
- Writing the report: Export coded segments, theme summaries, and analytic memos directly into word processing or reference management software, so you can trace every finding back to its source.
Where the Lumivero AI Assistant fits in
The NVivo AI Assistant is task-based—you select a specific action for it to perform, rather than working through open-ended chat—and it's designed to reduce repetitive work while you stay in control of interpretation. It shows up most during familiarization, coding, and theme refinement:
- Summarize a document or a selected passage, with a configurable summary length, to help you get oriented quickly during familiarization
- Get suggested child codes based on your existing coding, and run sentiment analysis to flag positive, negative, or mixed passages worth a closer look during coding
- Generate a draft summary within a Framework Matrix cell, synthesizing coded data for a given case and theme, during refinement
A few things worth knowing: summaries are generated one document at a time rather than as a single combined summary across multiple documents, and all AI-suggested codes and summaries can be reviewed, edited, or rejected. Organizations can also switch AI features on or off at the admin or local level—helpful for teams working under strict IRB, ethical, or data-privacy requirements.
Elevate your thematic analysis with NVivo
If you're ready to take your qualitative research to the next level, NVivo is your go-to tool for powerful, efficient thematic analysis. It helps you organize, code, and interpret large volumes of unstructured data—so you can move from raw transcripts to meaningful patterns with clarity and confidence.
