You have just finished your first round of fieldwork. Thirty interviews, each between 45 and 90 minutes, sitting in a folder on your laptop as raw transcripts. Your supervisor told you in your last meeting to "get them into NVivo." A colleague in the office next door swears by MAXQDA. Someone on the PhD forum you lurk on says they use a free tool called Taguette and it does everything they need. And you have seen a few Twitter threads about AI tools that apparently do the coding for you in minutes, though you are not sure whether that is legitimate or cheating.
This guide is for that moment. It covers what PhD-level qualitative analysis actually demands from a software tool, then goes through the main options including what they cost, what they cannot do, and what happens to your project files when your student discount expires three years from now. It ends with some practical advice about checking your institution's licences before spending any money.
What PhD-level qualitative analysis actually needs from a tool
Before comparing tools, it is worth being clear about what you actually need. Not every feature matters equally, and understanding your requirements will save you from paying for things you will never use.
Codebook development. Your analysis will likely produce a coding scheme that evolves over the life of the project. A good tool makes it easy to define codes, rename them, merge them, split them, and track how they changed over time. This is particularly important when your supervisor pushes back on your coding frame at your six-month review and you have to restructure from scratch.
Audit trail. Your methods chapter will need to describe your analytical process. A good tool records what you did and when, making it easier to write a credible account of your analytical journey. This is also relevant if you are ever asked to share your project files with a journal reviewer or your institution's data repository.
Export for appendices and sharing. You will likely need to include coded excerpts, codebooks, or frequency tables in your thesis appendices. You may also need to share your coding with your supervisor, who probably has their own preferred tool or none at all. Export formats matter more than they seem at the outset.
Longevity. A PhD typically takes three to five years. Whatever tool you choose, you need it to be available (and affordable) throughout that period. Student licences that expire with your student ID create a real problem if your writing-up takes longer than expected.
For a broader discussion of how to approach interview analysis at the methodological level, it is worth having that framing before getting into software choices.
NVivo student edition
NVivo is the dominant name in QDA software in most English-speaking academic contexts. If you are in management, social science, education, or health research, there is a good chance your supervisor trained on NVivo and considers it the default. Understanding what you are getting (and what you are not) is important before committing.
What you get
NVivo's feature set is genuinely broad. It handles text transcripts, PDFs, audio, video, images, and survey exports within a single project. Its query tools allow sophisticated cross-tabulations of codes against attributes (for example, comparing how a theme distributes across participants of different ages or roles). The visualisation tools, including word clouds, cluster analysis, and matrix coding queries, can surface patterns in large datasets.
NVivo also carries methodological credibility in citations. Writing "data were analysed using NVivo 15" is a recognised shorthand in methods sections; reviewers in many fields know what it means and accept it without question.
Pricing
NVivo is now owned by Lumivero, the private equity-backed company that also owns Atlas.ti. The student pricing for NVivo is approximately $118 (around €90–€110) per year for the academic student licence. This is significantly cheaper than the full academic licence, which runs to $295–$595 per year depending on tier.
Before purchasing, check whether your institution has a site licence. Many universities negotiate group agreements with Lumivero that give students free or subsidised access. Your library's software page or IT services team will be able to confirm. If your institution is covered, use that route before spending your own money.
The licence expiry problem
This is the issue that catches PhD students out more than any other. NVivo student licences are tied to your student status. When you graduate or your student ID expires, the licence expires with it.
This creates a practical problem: NVivo project files (.nvpx) are proprietary. You can export your data and codes, but the project environment itself, with all its coding relationships, queries, and memos, is not accessible without an active licence. If your thesis writing extends past your student period, or if you want to revisit the project for a follow-up paper two years after graduating, you will need to purchase a new licence or rebuild the project in another tool.
The workaround is to export your codebook, coded excerpts, and any reports you might need before your licence expires, and to keep copies of your original transcripts somewhere accessible. It is not a disaster, but it requires planning.
For a more detailed comparison of NVivo against its main competitors, including the full pricing picture, the NVivo vs MAXQDA guide covers this in depth.
The learning curve
NVivo rewards investment. Most students report spending several days, sometimes a week or more, learning the interface before they can work efficiently. If you are starting your analysis now and your supervisor wants to see results in six weeks, this is a real cost to factor in.
MAXQDA student edition
MAXQDA is made by VERBI GmbH, an independent German company that has been building qualitative software since the 1980s. Unlike NVivo, MAXQDA has no private equity ownership, and its pricing has been relatively more stable over time.
What you get
MAXQDA is consistently rated as easier to learn than NVivo. The side-by-side panel layout, showing your document list, coded segments, and codebook simultaneously, maps well to how qualitative researchers actually work. Most students reach productive coding within a day rather than a week.
Full Mac and Windows feature parity is a genuine differentiator. NVivo has historically offered fewer features on Mac; MAXQDA does not have this problem. If you work on a Mac, this alone can be a decisive factor.
For mixed methods theses, MAXQDA Analytics Pro includes statistical integration that lets you combine qualitative coding with quantitative analysis within a single project. This is particularly valuable for researchers using sequential or concurrent mixed methods designs.
Pricing
MAXQDA student pricing is broadly comparable to NVivo at the student tier, typically around €99 per year for the standard package. The Analytics Pro version (which includes mixed methods features) costs more. As with NVivo, check for institutional site licences before purchasing.
The mixed methods advantage
If your thesis combines qualitative interview data with survey results or other quantitative data, MAXQDA's integrated approach can save significant time. The alternative, managing qualitative coding in one tool and quantitative analysis in SPSS or R, is functional but involves a lot of data movement. For researchers in psychology, health sciences, or education research where mixed methods is common, this is worth considering seriously.
Atlas.ti student edition
Atlas.ti is the third major player in traditional QDA software. It was historically an independent competitor to NVivo, but it was acquired by Lumivero in 2023, making it now a sibling product to NVivo under the same private equity ownership. This is worth knowing when you see it positioned as an "alternative" to NVivo.
What you get
Atlas.ti has a strong following in grounded theory research and in social sciences outside the Anglo-American mainstream, particularly in German-speaking contexts and in South America. Its network view, which allows researchers to visualise relationships between codes, quotations, and memos as a graph, is a distinctive feature with no direct equivalent in NVivo.
Recent versions have added AI-assisted coding features, though these have received mixed reviews for the level of analytical control they provide.
Pricing
Atlas.ti student pricing is approximately $119 per year (pricing can vary by region). Like NVivo, the student licence is tied to student status, and the same expiry considerations apply.
Given that Atlas.ti and NVivo now share a parent company, the competitive distinction between them is less pronounced than it once was. If your supervisor or field has a specific preference for Atlas.ti, that is a valid reason to choose it. Otherwise, the choice between Atlas.ti and NVivo is less meaningful than the choice between Lumivero tools and independent alternatives.
For a broader view of how all these tools compare, including newer AI-native alternatives, the complete QDA tools comparison covers the full landscape.
Free options: Taguette and QualCoder
Before spending money on any of the above, it is worth understanding what the free tools can and cannot do.
Taguette
Taguette was created at NYU Tandon with an explicitly equity-driven motivation: to give researchers without large research funds access to workable QDA software. It is free, open source, and runs either locally on your machine or via a hosted browser interface.
For straightforward document highlighting and tagging, Taguette works well. You can import transcripts, highlight passages, assign codes, and export tagged excerpts. For a small qualitative study (under 20-30 documents with a manageable number of codes), it can handle a full thesis analysis.
The limitations are real at PhD scale. Taguette does not support hierarchical codes, so if your codebook has multiple levels of abstraction, you will need to manage that in a separate spreadsheet. It does not support multimedia files or advanced queries. For a thesis that needs to demonstrate methodological rigour in the ways described above, these gaps matter.
QualCoder
QualCoder is a more fully featured open-source option. It supports hierarchical coding and a wider range of file types than Taguette. It is free to download and run locally.
The interface is more utilitarian than either NVivo or MAXQDA, and the documentation is thinner. For technically confident researchers who are comfortable working with software that requires occasional troubleshooting, QualCoder is a legitimate option. For researchers who want polished software with professional support, it will feel rough.
When free tools are enough
If your study is small in scope (under 30 interviews, relatively flat code structure, no multimedia) and you have a patient and technically capable approach to software, Taguette or QualCoder can handle the work. The University of Oregon Libraries, which stopped subscribing to commercial QDA software in 2024, now recommends free tools as sufficient for many research needs.
For larger or more methodologically demanding theses, the feature gaps will catch up with you.
AI-native tools: Skimle
The newest category in qualitative research software is tools built around AI from the ground up. NVivo and MAXQDA have added AI features to existing architectures, but they remain fundamentally manual coding environments with AI sprinkled on top. AI-native tools work differently.
Skimle is designed so that AI performs the initial coding pass across your entire dataset, building a structured representation of themes and coded excerpts. The researcher then reviews, edits, challenges, and extends that structure. This is not the same as asking ChatGPT to summarise your interviews. The AI reads all your documents systematically, builds a codebook bottom-up from the data, and links every theme back to the specific paragraphs that support it.
Where this matters for PhD research
For a dataset of 30 interviews, AI-assisted initial coding can reduce the time to first-pass analysis from weeks to days. This does not mean the intellectual work of analysis disappears. You still have to decide what the themes mean theoretically, how they relate to each other, and how they connect to existing literature. But the mechanical first pass of working through transcripts and assigning codes is where AI creates the most obvious time saving.
Manual editing control. Skimle allows full manual editing of AI-generated categories: add, delete, relabel, merge, and regroup codes, move individual excerpts between themes, and manually code passages the AI missed or miscategorised. You retain the same interpretive control you would have in NVivo or MAXQDA, with significantly less time spent on the initial pass. This matters for thematic analysis done properly, where the researcher's interpretive judgement is the core of the method.
REFI-QDA export. If your supervisor works in NVivo or your co-author uses Atlas.ti, Skimle exports in the REFI-QDA (.qdpx) standard. This open format is supported by NVivo, Atlas.ti, and MAXQDA, meaning you can run initial analysis in Skimle and hand off to a colleague's preferred tool without losing coding work. It also gives you a format for long-term data archiving that does not depend on Skimle remaining your active tool.
Codebook export to Word. Your supervisor almost certainly does not have a Skimle account. Skimle exports the codebook as a Word document with each coded excerpt marked as a comment in the margin. This is a format any supervisor can review without needing to install or learn new software.
Audit trail and transparency. Every theme in Skimle links back to the exact source paragraphs, and the AI's reasoning is visible, not hidden. For a methods chapter that needs to describe and defend the analytical process, this two-way traceability is directly useful. You can show exactly which passages support which themes, which is what a rigorous methods section requires.
AI disclosure in academic methodology
The growing acceptance of AI in qualitative research comes with an expectation of transparency. Using AI in qualitative research is increasingly accepted in academic publishing, but the consensus is that methodology sections must disclose the tool used and the role it played. "Interviews were analysed using Skimle's AI-assisted thematic analysis, with full manual review and editing of all AI-generated codes" is the kind of disclosure that is becoming standard. The parallel with NVivo, where researchers do not normally agonise over whether it is legitimate to use software for qualitative analysis, is a reasonable frame: the question is about transparency and rigour, not about whether AI assistance invalidates the work.
For a fuller discussion of methodological transparency in AI-assisted research, including the specific questions journal reviewers are likely to ask, the academic researchers' AI guide covers this directly.
Sample size and what your tool needs to handle
One consideration that PhD students sometimes underestimate when choosing tools is how many documents they will end up with. A study that starts as 20 interviews often grows. Your supervisor suggests adding some document analysis. You decide to include secondary interview data from a related project. The appendices require more coded excerpts than you expected.
Qualitative research sample sizes in top management journals have increased considerably over the past decade. Studies with 50 or more interviews are now common in journals like Academy of Management Journal. If your thesis aims at journal publication after graduation, the dataset you build now may need to scale further than your initial plan suggests.
Manual tools handle this if you have the time. AI-native tools handle it with less friction. Free tools start to creak at scale.
Practical advice before you start
Check your institution's site licences first
Before purchasing anything, contact your university library or IT services. Many institutions negotiate site licences with NVivo, MAXQDA, or both that give students free or heavily subsidised access. Some institutions have recently dropped these licences (NVivo lost a major UK academic consortium deal in July 2024), so it is worth checking the current status rather than assuming access exists.
Understand what format your supervisor expects
If your supervisor expects to see your NVivo project files, buying MAXQDA creates a compatibility problem. If your supervisor is agnostic, you have more flexibility. Have this conversation early, ideally in your first few months of fieldwork planning, not after you have already coded 25 interviews.
Plan for what happens when your student licence expires
Whatever tool you choose, make a plan for data export before your student status ends. Export your codebook, your coded excerpts, your memos, and your original transcripts in formats that do not require the original software to open. This is basic data management hygiene, but it is easy to forget until it becomes urgent.
Think about how to work effectively with interview transcripts
The best tool is the one that fits your analytical process. If you already have a clear sense of how you work with data, choose the tool that scaffolds that process. If you are less certain, starting with a free tool for your first few interviews before committing to a paid licence is a reasonable approach.
Summary: which tool fits which situation
To bring this to a practical conclusion:
Use NVivo if your institution has a site licence, your dataset includes multimedia, your field has strong NVivo norms, and you are on Windows. The learning curve is steep but the feature set is the broadest available in traditional QDA software.
Use MAXQDA if you are paying yourself and want the best value for a full-featured tool, you work on a Mac, or your thesis combines qualitative and quantitative analysis. It is generally easier to learn than NVivo and the independent ownership is a minor but real advantage.
Use Atlas.ti if you have a specific reason related to your field or supervisor preference. Its analytical heritage in grounded theory is genuine, but its acquisition by Lumivero removes the competitive independence argument it previously had.
Use Taguette or QualCoder if your project is small in scope, your coding needs are relatively straightforward, and cost is a hard constraint.
Use Skimle if you want to recover the weeks typically spent on initial coding, need REFI-QDA export for compatibility with supervisor or co-author tools, want a codebook you can share with a supervisor in Word format, and value an audit trail where every theme links directly to source evidence. For researchers interested in a practical overview of what AI-native analysis offers versus manual approaches, that framing helps place the tool choice in a methodological context.
Ready to cut weeks from your initial coding pass and still have full control over every category and code? Try Skimle for free and run AI-assisted thematic analysis on your interview transcripts, with a full audit trail and codebook export your supervisor can review in Word.
Want to go deeper on methodology? Read our guides on how to use AI in qualitative research for academics, thematic analysis from start to finish, and REFI-QDA export and manual coding workflows.
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
- NVivo pricing — Lumivero
- MAXQDA pricing — VERBI GmbH
- Atlas.ti student pricing — Atlas.ti
- Lumivero acquires ATLAS.ti
- NVivo changes via Chest UK consortium — July 2024
- Taguette — official site
- QualCoder — GitHub
- University of Oregon Libraries — free qualitative data analysis software
- REFI-QDA standard — KWALON/QDASR
- ResearchGate discussion: NVivo vs Atlas.ti vs MAXQDA (2023)