This week’s guest blogpost is authored by Dr. Matthew J. Smith. Matthew is a lecturer in the program of Higher Education Leadership in the Department of Leadership, Technology, and Workforce Development at Valdosta State University in Georgia. He has research interests in C]college student learning and development outside of the formal classroom setting; sense of belonging; graduate students; professional school students (particularly those in the health sciences); assessment practices; leadership development; social class; research design (qualitative, quantitative, and mixed-methods); and humanities education (particularly history).
In qualitative research, interviews are a common way of collecting qualitative data, particularly since interviews are frequently how we as a society gather knowledge on the range of things and events happening in the world (Roulston, 2022). Recent scholarship postulates to qualitative scholars to consider both current challenges and future opportunities with transcription in qualitative research (Azevedo et al, 2017; Dobbs et al., 2024; Eftekhari, 2024; Jenkins et al., 2023; McMullin, 2023; Point & Baruch, 2023). One often overlooked ethical benefit of AI-facilitated transcription is removing the possible negative impact of transcription of interviews on sensitive topics may have on the transcriptionist (Hennessy et al., 2022). Given this ethical benefit, it would behoove qualitative scholars to actively consider utilizing AI-facilitated transcription options for their work.
Scholars have performed studies comparing transcriptions completed using various platforms (Siegel, et al., 2023; Sponholz et al, 2025; Wollin-Giering et al., 2024), while others have sought to develop more accurate AI-based transcription/analysis options (Tolle et al, 2024). While doing data collection for a study using primarily interviews for data, I elected to compare human transcription (specifically Rev.com), verbatim AI transcription (specifically Rev.com) and multiple AI transcription options without verbatim (specifically Microsoft Teams, Rev.com AI, All options were in line with IRB approval for transcription of interviews for this study. Below I will briefly discuss my perceived pros and cons of each.
Rev.com (Human Transcription)
Even with the proliferation of AI-facilitated transcription options, there are still places one can obtain transcription completed by a person. Due to the relative widespread use of Rev.com, I elected to utilize this platform for this comparison. The transcriptions are typically completed within one day, and are advertised as being 99% accurate. Based on my experience with this comparison (and prior experience using this particular service with Rev.com in other research projects), their statement on accuracy seems quite apt. The errors were few, and any place in the transcript there is a portion they are unsure of, the transcription has the time flagged and noted in such a way that it is easy to examine as the researcher.
While there are significant benefits to this transcription service, it does come with some drawbacks compared to the others. The first is the significantly higher cost. At the time I completed this ad hoc comparison, Rev.com was $1.99 per minute for human transcription. For long and/or multiple interviews, this can get costly quickly. Another potential drawback is the time needed for the file to be transcribed. The AI and automatic transcriptions were all completed in a matter of minutes, while the human transcription took close to a day to complete. While still swift, if time is a pressing concern to the researcher, this comparative delay for transcription could be an issue.
Rev.com (With and Without Verbatim)
Rev.com offers AI-facilitated transcription (both with and without verbatim of the transcription). There is no additional cost between the two options, so I am discussing them jointly here. Compared to the human transcription offered by Rev.com, the cost is MUCH lower using this service ($0.25 per minute at the time of this study). Furthermore, the transcripts are returned in a matter of minutes (often less than 10 minutes for transcripts of interviews up to an hour in length). The accuracy is nearly identical to the transcription done by a human with Rev.com, and is much more accurate than the transcription provided by Microsoft Teams.
In terms of cons, it is worth noting that the accuracy is slightly less than that of human transcription. If accuracy is of paramount importance to a researcher, they may want to consider a non AI-facilitated transcription option. While it is lower cost than the human transcription options, it still has a cost compared to other AI-facilitated transcriptions, which is a point to be mindful of when we as researchers frequently have to work with limited resources. It is worth noting that the AI-facilitated transcripts from Rev.com seem to flag places where it is unsure of a word or phrase much less often than human transcription. Given this, a closer listening of the transcript for verification is needed to ensure accuracy.
Microsoft Teams
Microsoft Teams provides several benefits in terms of AI-facilitated transcription. First and foremost is that it is free when you are using Microsoft Teams. Other platforms such as Zoom offer this as well as part of utilizing their software. Given these kinds of software are included in the computer options given to scholars and students at many universities, the “free” cost for this can make it a very attractive option. This may be particularly attractive for scholars with limited supports and means (Isangula, 2025). Microsoft Teams does the transcription as you conduct the interview, allowing you to have access to the raw transcript essentially as soon as you conclude the interview. This can allow for cleaning and preliminary analysis to occur quickly if that is desirable to the researcher. One drawback of Microsoft Teams was there were (in my experience) frequent errors with the transcription. Some were just small mistakes with words, but other cases it coded entire sentences and lines under the wrong speaker. This can make for a more involved cleaning and verification process for the researcher(s), possibly prolonging the process due to these frequent issues.
Recommendations and Conclusion
While the unique circumstances of each researcher may vary, in my opinion the verbatim AI-facilitated transcription offered by Rev.com offers the best balance of accuracy, speed of transcription, accessibility, and affordability. If time and finances are not limited resources (a luxury few scholars have), I would fully endorse the human transcription service offered by Rev.com. Considering the reality many scholars and researchers face, the AI-facilitated verbatim transcription offered by Rev.com provides the best mix of accuracy, speed, and utility for a reasonable price.
Authors note: I am not being compensated by any of the services I used as part of this comparison. This is solely based on my own experiences. All thoughts and opinions (unless otherwise noted) are my own.
References
Azevedo, V., Carvalho, M., Fernandes-Costa, F., Mesquita, S., Soares, J., Teixeira, F., & Maia, A. (2024). Interview transcription: Conceptual issues, practical guidelines, and challenges. Revista de Enfermagem Referência, IV(14), 159-168.
Dobbs, D., Yauk, J., Meng, H., & Jalil, T. (2024). The use of artificial intelligence transcription technology in caregiver qualitative data collection. Innovation in Aging, 8(S1), Article 459.
Eftekhari, H. (2024). Transcribing in the digital age: Qualitative research practices utilizing intelligent speech recognition technology. European Journal of Cardiovascular Nursing, 23(5), 553-560.
Hennessy, M., Dennehy, R., Doherty, J., & O’Donoghue, K. (2022). Outsourcing transcription: Extending ethical considerations in qualitative research. Qualitative Health Research, 32(7), 1197-1204.
Isangula, K. G. (2025). Navigating barriers: Challenges and strategies for adopting artificial intelligence in qualitative research in low-income African context. Tanzania Journal of Health Research, 26(3), 2048-2059.
Jenkins, N., Monaghan, K., & Smith, M. (2023). Did they really say that? An agential realist approach to using computer assisted transcription software in qualitative data analysis. International Journal of Social Research Methodology, 26(1), 97-109.
McMullin, C. (2023). Transcription and qualitative methods: Implications for third sector research. Voluntas, 34(1), 140-153.
Point, S., & Baruch, Y. (2023). (Re)thinking transcription strategies: Current challenges and future research directions. Scandinavian Journal of Management, 39, Article 101272.
Roulston, K. (2022). Interviewing: Guide to theory and practice. SAGE Publications, Inc.
Siegel, R., Mrowczynski, R., Hellenthal, M., & Schilling, M. (2023). Poster: From hashes to ashes – A comparison of transcription services. Proceedings of the 2023 ACM SIGSAC Conference on Computer and Communications Security, 3534-3536. https://doi.org/10.1145/3576915.3624380
Sponholz, J., Weilinghoff, A., & Schopf, J. (2025). Halving transcription time: a fast, user-friendly and GDPR-compliant workflow to create AI-assisted transcripts for content analysis. arXiv, 2503, Article 13031. https://doi.org/10.48550/arXiv.2503.13031
Tolle, H., del Mar Castro, M., Wachinger, J., Putri, A. Z., Kempf, D., Dekinger, C. M., & McMahon, S. A. (2024). From voice to ink (Vink): Development and assessment of an automated, free-of-charge transcription too. BMC Research Notes, 17, Article 95.
Wollin-Giering, S., Hoffmann, M., Hӧfting, J., & Ventzke, C. (2024). Automatic transcription of English and German qualitative interviews. Forum: Qualitative Social Research, 25(1), Article 8.
QualPage 