Qualitative data is any information that takes the form of words, images, sounds, or observed behaviour rather than numbers. Common types include interview transcripts, focus group recordings, open-ended survey responses, field notes, policy documents, and social media text. Each type captures something different about human experience and is suited to different research questions.
This article covers eight types of qualitative data with examples from academic research and business practice. For a broader overview of what qualitative data is and how it differs from quantitative data, see what is qualitative data.
What counts as qualitative data?
The clearest way to define qualitative data is by what it preserves: context, nuance, meaning, and language. A qualitative data point is not a number on a scale; it is a sentence, a story, a gesture, or a document that conveys something about how people think, feel, or behave.
Qualitative data is collected when:
- The research question asks "how," "why," or "what does this mean" rather than "how many" or "how much"
- The phenomenon being studied is too complex, contextual, or underexplored to reduce to a pre-designed scale
- The researcher needs to understand the range of experiences or views rather than estimate prevalence
The eight types below vary considerably in how they are collected, how they are structured, and what kind of analysis they support.
The 8 main types of qualitative data with examples
1. Interview transcripts
What they are: A word-for-word written record of a spoken conversation between a researcher and one or more participants. Can be structured (fixed questions in fixed order), semi-structured (topic guide with flexibility), or unstructured (exploratory conversation).
Academic example: A sociologist studying the experience of long-term unemployment conducts 24 semi-structured interviews with adults who have been unemployed for more than 12 months. Each interview lasts 60-90 minutes. The transcripts, totalling around 150,000 words, are then analysed using thematic analysis, producing themes around identity disruption, social withdrawal, and systemic frustration with job market structures.
Business example: A B2B software company runs 15 customer discovery interviews before building a new product feature. Each 45-minute interview explores the participant's current workflow, pain points, and workarounds. The transcripts are coded for recurring problems and "job to be done" moments.
Strength: High depth of insight per participant; allows follow-up questions; captures exact language used. See how to analyse interview transcripts for the full analytical process.
Limitation: Time-intensive to conduct and transcribe. Each interview yields a lot of raw text that requires systematic coding.
2. Focus group transcripts
What they are: A transcript of a facilitated group conversation (typically 6-10 participants) on a defined topic. The group dynamic produces data that individual interviews cannot: disagreement, consensus-building, social pressure, and spontaneous elaboration.
Academic example: A health psychologist convenes five focus groups with parents of young children to explore attitudes toward childhood vaccination. Each group is 90 minutes. The transcripts are analysed to identify how social norms and media exposure shape vaccination decisions, with particular attention to how participants modify their views in response to others in the group.
Business example: A consumer goods brand runs six focus groups across three European markets to explore attitudes toward a proposed reformulation of a food product. Transcripts reveal that the emotional significance of the original recipe ("it tastes like childhood") is far stronger than anticipated, a finding that changes the brand's communication strategy.
Strength: Group dynamics surface norms and social context; efficient for exploring shared or contested views across a population.
Limitation: Dominant voices can suppress minority views; findings are context-dependent (what people say in a group is not necessarily what they believe alone). For a comparison of this method with individual interviews, see focus groups vs individual interviews.
3. Field notes and observation records
What they are: A researcher's written record of what they observed in a naturalistic setting. Field notes typically include descriptive observation (what happened, who was present, what was said), reflective observation (the researcher's interpretations and questions), and methodological notes (decisions made about data collection).
Academic example: An organisational ethnographer spends eight months working alongside employees at a logistics company undergoing digital transformation. Field notes document informal conversations, how workers respond to the new systems, workarounds they develop, and the gap between management's stated rationale and employees' experienced reality.
Business example: A UX research team conducts contextual inquiry at a hospital, observing how nurses use a new software system during their shifts. Field notes record every point at which a nurse pauses, asks a colleague for help, or reverts to paper. These observations feed directly into the product redesign brief.
Strength: Captures actual behaviour in context rather than recalled or reported behaviour; surfaces tacit knowledge that participants cannot articulate.
Limitation: Labour-intensive; data collection is constrained to what the researcher can observe; the observer's presence can affect what happens (Hawthorne effect).
4. Open-ended survey responses
What they are: Text responses to open-ended questions within a survey instrument. Often combined with quantitative (Likert scale or multiple choice) questions in the same survey.
Academic example: A political scientist includes two open-ended items in a 40,000-respondent survey on attitudes toward climate policy: "In your own words, what concerns you most about climate change?" and "What, if anything, would make you more supportive of carbon pricing?" The responses are analysed using content analysis, producing frequency counts for broad categories and qualitative analysis of distinctive framings.
Business example: A SaaS company with 12,000 customers sends an annual NPS survey with the standard "How likely are you to recommend?" question followed by an open-ended "Why did you give that score?" question. The 4,000 verbatim responses are coded by theme (onboarding, support quality, feature gaps, pricing, competitor comparison) and cross-tabulated by NPS segment. For guidance on this kind of analysis, see how to analyse open text responses at scale.
Strength: Can reach very large samples; quantitative and qualitative data collected in the same instrument.
Limitation: Responses tend to be shorter than interview data; context is absent; the question frames the response.
5. Documents, reports and archival materials
What they are: Any written record produced for purposes other than research: policy documents, annual reports, internal memos, court transcripts, newspaper articles, historical archives, regulatory submissions.
Academic example: A historian studying the development of environmental regulation in the UK analyses 40 years of parliamentary debates, government reports, and industry submissions. The analysis traces how the framing of environmental risk shifted from a technical question (measurable pollution levels) to a political question (whose interests should be protected) over the period.
Business example: A private equity team conducting commercial due diligence analyses 60 regulatory filings, 25 earnings call transcripts, and 30 customer-facing documents from a target company's competitors. The qualitative analysis identifies how competitors are positioning on compliance burden and what language they use to describe regulatory risk to investors. See earnings call transcript analysis with AI for more on this application.
Strength: Unobtrusive data collection (no observer effect); historical depth; access to perspectives that are not available for interview.
Limitation: Documents reflect the intentions and perspectives of their authors; what is not written down is invisible; access can be restricted.
6. Social media and online text
What they are: User-generated text from social media platforms, review sites, forums, blogs, and comment sections. Produced by people for their own purposes rather than in response to a researcher's questions.
Academic example: A communication researcher analyses 10,000 Reddit posts from a community discussing mental health, using discourse analysis to examine how members negotiate the boundary between professional help-seeking and peer support.
Business example: A consumer electronics company analyses 50,000 app store reviews across three product lines, coding each review for themes (battery life, camera quality, software stability, customer service) and sentiment. The analysis is used to prioritise the product roadmap and identify which issues are driving one-star reviews.
Strength: Naturally occurring data; very large volumes available; captures perspectives that people would not share with a researcher directly.
Limitation: Context is limited; anonymity means demographics are unknown; content moderation creates selection effects; ethical questions around consent for research use.
7. Audio and video recordings
What they are: Recordings of natural interaction, interviews, presentations, or observed behaviour. Video adds visual context (gesture, facial expression, physical environment) to spoken content.
Academic example: A conversation analyst records and transcribes 30 medical consultations, analysing the precise sequence of turn-taking between doctors and patients to understand how diagnoses are delivered and received. The analysis attends to pauses, overlapping speech, and the moment at which the patient shifts from passive recipient to active questioner.
Business example: A contact centre analyses recordings of 500 customer service calls using a combination of automated transcription and qualitative coding. The qualitative analysis identifies the specific phrases agents use when calls escalate to complaints, and develops a training programme based on what differentiates calls that de-escalate successfully from those that do not.
Strength: Captures non-verbal communication; provides a permanent record that multiple researchers can analyse; supports very detailed sequential analysis.
Limitation: Transcription is time-consuming; video data raises privacy concerns; participants are aware of being recorded.
8. Research diaries and participant journals
What they are: Written or audio-recorded accounts produced by participants over time, documenting their experiences, thoughts, and behaviours. Often used in longitudinal research or where the researcher cannot be present.
Academic example: A UX researcher recruits 20 participants for a two-week diary study of how they use their smartphones during commutes. Participants submit a voice memo at the end of each commute describing what they did, why, and how they felt about it. The diary data reveals temporal patterns and incidental use that a single interview would not capture.
Business example: A customer experience team asks 30 customers to keep a journey diary during their first 90 days as subscribers, recording any point at which they feel confused, frustrated, or delighted. The diaries are coded to identify the precise moments in the onboarding experience that make or break long-term retention.
Strength: Captures change over time; reduces retrospective bias (participants record experiences while they are fresh); reaches moments the researcher cannot observe.
Limitation: Participant burden is high and compliance drops over time; entries vary in length and detail; requires careful relationship management to sustain engagement.
How do these data types compare?
| Data type | Depth per item | Volume potential | Context richness | Analysis effort |
|---|---|---|---|---|
| Interview transcripts | Very high | Low-medium | High | High |
| Focus group transcripts | High | Low | High | High |
| Field notes | High | Low | Very high | High |
| Open-ended survey responses | Low-medium | Very high | Low | Medium |
| Documents | Variable | High | Variable | Medium |
| Social media text | Low | Very high | Low | Medium |
| Audio/video recordings | Very high | Low | Very high | Very high |
| Research diaries | Medium-high | Low | Medium | Medium |
The right data type depends on the research question, access, resources, and the kind of analysis you plan to conduct. Most real research projects combine several types.
What software do researchers use to analyse qualitative data?
The main options for qualitative data analysis (QDA) are:
Traditional QDA tools: NVivo, MAXQDA, and ATLAS.ti are designed for manual coding of text and multimedia data. They are comprehensive but expensive and have a steep learning curve. See qualitative data analysis tools comparison for a full breakdown.
AI-assisted tools: Skimle processes text-based qualitative data (transcripts, documents, survey responses) through a structured AI analysis pipeline that codes, categorises, and synthesises findings while maintaining traceability from every insight back to the source. How Skimle works explains the approach in detail.
Spreadsheet-based methods: Many practitioners use Excel or Google Sheets for simpler coding tasks, particularly with open-ended survey data. This works for small datasets but becomes unwieldy at scale. See how to do thematic analysis in Excel for a step-by-step guide.
Frequently asked questions
Is video footage qualitative data?
Yes. Video recordings are qualitative data. The content (what people say and do) is analysed through interpretation, not measurement. When researchers analyse video, they may transcribe speech, describe observed behaviour, and interpret gestures and facial expressions. The analysis is qualitative even if the output includes some frequency counts.
Can qualitative data be converted into quantitative data?
Yes, through coding and quantification. When you code qualitative data into categories and count how often each category appears, you are quantifying qualitative data. This is the basis of content analysis and is also used in mixed methods research to integrate qualitative and quantitative findings. For more on this, see how to quantify qualitative data.
What is the difference between qualitative data and qualitative research?
Qualitative data is the material (transcripts, field notes, documents). Qualitative research is the systematic process of collecting and analysing that data using appropriate methods (thematic analysis, grounded theory, discourse analysis, etc.) to answer a research question. You can have qualitative data without doing qualitative research (a collection of interview transcripts is just text until it is analysed), but you cannot do qualitative research without qualitative data.
How do you ensure quality when using qualitative data?
Quality in qualitative research is assessed through criteria like credibility (are the findings grounded in the data?), transferability (how far do they apply beyond the sample?), dependability (could another researcher follow your analytical process?), and confirmability (can you demonstrate that your interpretations are anchored in data rather than researcher assumptions?). Practical approaches include member checking, audit trails, thick description, and negative case analysis. See the AI qualitative data analysis checklist for a practical pre-publication quality framework.
Working with qualitative data and looking for a faster, more systematic way to analyse it? Try Skimle for free. Upload transcripts, documents, or survey responses and see AI-assisted coding and categorisation with full traceability from every insight back to its source.
Related reading: What is qualitative data? Types, examples and how to analyse it | How to analyse interview transcripts | How to analyse open text responses at scale
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
Sources
- Qualitative Research and Evaluation Methods — Patton (2014), SAGE
- Basics of Qualitative Research — Strauss & Corbin (2014), SAGE
- Doing Survey Research — Bryman (2016), SAGE
- Interaction in the Language Classroom — Seedhouse (2004), Pearson
- Experience Sampling Method — Csikszentmihalyi & Larson (1987), Journal of Nervous and Mental Disease
