From interview to traceable findings
Qualitative analysis is not linear; you move from coding to themes and back again. The panel supports that movement and records every step.
1. Bring your data
5Three routes: file upload, pasted text, survey import.
- Word (.docx), PDF with a text layer, plain text (.txt, .md), subtitles (.vtt, .srt) and CSV; subtitle and CSV files are read as plain text
- Automatic splitting of speaker turns written as “Name:”
- Speaker → participant mapping for focus groups
- Open-ended answer import from YouReply Survey
- Participant attribute import from CSV
2. Code
4The core of the product: fully usable without any AI.
- Assign codes by highlighting text
- Overlapping and nested codings
- Hierarchical codebook with definitions and rules
- Excerpt and PARTICIPANT counts per code
3. Optional: Review AI suggestions
4Optionally request code suggestions against your existing codebook or ask for new code candidates. Review every suggestion with its source, then accept, edit, or reject it. Manual coding, the codebook, themes, matrices, memos, and reports remain available when AI is disabled.
- Deductive coding against your codebook
- Inductive coding with new code proposals
- Suggestion inbox: accept / change code / reject
- A suggestion that does not match the source text cannot be accepted
4. Consolidate the codebook
5The second round inflates the codebook; tidying it is part of analysis.
- Merge, split, move, archive, rename
- See how many codings a merge will affect before you confirm
- Rationale note for merges and splits
- Undo the last operation in one step
- AI-suggested merge groups
5. Build themes
4A theme is not a parent code; it is an interpretive layer above codes.
- A theme may draw codes from different branches
- A code may belong to several themes
- Theme narrative: the body text of your report
- Prevalence is computed over PARTICIPANTS
6. Compare
4Matrices and patterns; a matrix cell opens the excerpts behind it.
- Theme × participant matrix
- Code × demographics matrix
- Similarity, difference, contradiction and outlier suggestions
- Findings stay out of reports until accepted
7. Write memos
3In grounded theory, the analytic memo IS the analysis.
- Analytic, methodological, reflexive and general memos
- Attach to a document, code, theme or excerpt
- Search excerpts, memos and uncoded transcript text
8. Report
4Codebook, quote table, matrix and full report, all generated from your coded data.
- Codebook report
- Quote table with sources
- Matrix report and full qualitative research report
- Word and Excel download, PDF from your browser
Tracing a finding back to its source
In the workspace the link between theme, code, excerpt and source document is built in this order. The chain below uses fictional sample data.
Fictional sample data · Illustrative view
1. Theme
Blurred boundaries in flexible work
The researcher's interpretive layer. It draws on several codes; the theme narrative is written by the researcher.
2. Code
Boundary between work and home
Definition: passages where the participant describes difficulty separating working hours from private life. Exclude: passages that only mention commuting time.
3. Excerpt
“I can't even put my phone away at dinner.”
Participant P3 · Origin: AI suggestion, accepted by the researcher.
4. Source document
Interview 3 (transcript.docx)
The excerpt is stored with its character range in the document; the quote text in the report is taken from that range.
What the counts mean
Participant, excerpt, and code counts help researchers navigate the analysis, compare patterns, and identify divergent views. These indicators do not by themselves establish statistical representativeness, significance, or generalizability to a population. Qualitative findings should be evaluated together with context, research design, and researcher interpretation.
The researcher's responsibility
YouReply Qualitative supports coding, theme development, source tracing, and draft reporting; it does not replace qualitative research advice. The researcher should verify the codebook, themes, quotes, interpretations, and final report.
Why you can defend these findings
A number you cannot defend is useless in an academic study.
Every finding is anchored
The chain theme → code → excerpt → character range in the document → participant is never broken. You can always show where a quote came from.
The AI never finalizes alone
An unaccepted suggestion never enters a matrix, a count or a report. How much of the coding started as an AI suggestion is recorded and can be declared in your paper.
Quotes remain linked to their source
AI suggestions cannot be accepted unless they are linked to a matching passage in the source text; matching ignores differences in letter case, whitespace and quotation marks. Quotes used in reports are retrieved from the stored source rather than rewritten by the model. Suggestions whose source cannot be verified are rejected or flagged for review. Narrative text written by the AI, such as theme descriptions and report drafts, is not a quote. The researcher should still verify every quoted passage and its context before reporting it.
You can switch it off
AI can be disabled for an individual study; when disabled, no new AI analysis is started for that study and manual coding features remain available. When a new study is created this setting currently starts switched on; if you will not use AI, turn it off while creating the study. Turning AI off does not by itself satisfy ethics committee or data protection obligations: This setting alone does not satisfy ethics approval, participant notice, consent, legal-basis, or institutional-policy requirements.
Code your first interview today
The free plan carries a pilot study from start to finish. No credit card required.