☀️ Happy Sunday!
Greetings from a chilly New York. How are you?
I’m excited to invite you to Swarna, an amazing user researcher from our Design with AI community. We first met at an ADPList’s Meetup in Central Park, and I was surprised to learn that she, too, had an architecture background before pivoting to UX. Today, she will be sharing her insights on Dovetail, an AI-powered platform in UX research.
While I use UserTesting often for user research, I’ve been hearing a lot about Dovetail recently. First, a student from my AI course mentioned she enjoyed using it. Then Swarna told me she found Dovetail helpful too. I was curious to learn more.
So here we are. Please join me in welcoming Swarna!
Hello!
As an architect turned UX researcher, one thing I’ve consistently loved across both fields is interacting with users. Engaging with people before designing physical spaces was a highlight, and I continue to enjoy connecting with users to shape digital products today. What I didn’t love, though, was the documentation part—it felt like a tedious task, and I’d often wish I had an assistant to handle it for me.
The role of AI in our workflows is often controversial—it undoubtedly simplifies tasks but can feel threatening as well. Aren’t we all daunted by the thought of AI replacing our jobs? But what if AI can make our work more effective? I’d like to take this opportunity to highlight the positive impact AI has had on my research practice and share how it has enhanced my work.
My user research journey began with documenting data the old-school way—using traditional Excel sheets and Word docs. While Excel sheets have their perks, managing data this way becomes cumbersome. Transferring verbatim from transcripts, jotting down notes during user sessions, and then tagging and coding everything was incredibly time-consuming.
It wasn’t until early 2021, when a colleague introduced me to Dovetail, that my workflow took a turn for the better. The tool allowed me to organize my research data in a streamlined, structured way, making thematic analysis far more efficient and it served as a research repository. Yes, it's a paid tool, but if your organization provides access, it's a huge win. And if you don’t have it in your organization yet, hopefully this post will motivate you to request it—you’ll be glad you did!
What is dovetail?
For those unfamiliar with Dovetail, it's a comprehensive user research tool that allows you to generate transcripts from research videos or audio, tag and highlight key insights, and organize references. You can also use it to create detailed user research reports, making it an all-in-one platform for streamlining the research process.
How is AI in dovetail helping user research?
Dovetail recently got better with AI. About 3 months back, I started using the new AI features and here are my top 4 favorite:
1. Automatic transcription
When you create a project in Dovetail, you can add notes, and within each note, upload video or audio files. With Dovetail’s new feature, transcription kicks off automatically, and it’s done in under 2 minutes—something that I would take hours to transcribe manually.
2. Highlights feature
This feature uses AI to highlight the most important sections of the transcript, making the tagging process quicker for each one (Highlighted sections in Fig. 2). However, the researcher still needs to review the transcript to ensure that no key observation or theme is overlooked.
3. Clustering tags
Previously, I would do all my coding and theme tagging manually on Excel. With Dovetail, I can now create tags and themes directly within transcripts. All tags show up on a tag board (see Fig. 3), where I can also group them further. For clustering, tags can be found in the highlights section in the left navigation (see Fig. 3).
Clustering can be done by tags, themes, or notes (See Fig. 4). Essentially, this feature groups all text with the same tag and helps in further analyzing the data to derive insights.
4. Summary feature
This feature gives a quick summary (section on the right in Fig. 5). I find this feature very helpful in debriefs right after user interviews. It’s useful for an initial overview—however, take it with a pinch of salt, it still needs the researcher’s review, as the summary may not always be ready for use in the AI-generated format.
I recently discovered that Dovetail released its new version- Dovetail 3.0, which includes some cool new AI features like survey analysis and an “Ask Dovetail” option. I haven’t had a chance to try it out yet, but I’m excited to dive in. If you’ve already used it, please share your thoughts in the comments!
As UX practitioners, finding efficient ways to manage our time and focus on the human side of our work is essential, and that’s exactly where tools like Dovetail, powered by AI, step in. While it’s natural to feel wary of AI advancements, my experience with Dovetail has shown that these tools can significantly reduce tedious work. As we navigate these evolving tools, we can harness AI to elevate our research practices and continue focusing on what truly matters: understanding and advocating for our users. I'm eager to hear from other researchers about their experiences with AI-powered tools. If you’ve used Dovetail or other tools, kindly share what you like and don't like about these tools below!
Lastly, thank you very much Xinran for the opportunity to feature in your blog! More power to you for your initiative of empowering design professionals with the potential of AI.
Thanks Swarna for sharing! You can find her on LinkedIn.
What I like about the interface is that it shows the AI-generated info and the original transcript side by side (instead of mixing them together), making it easier to cross-check. The highlight feature is interesting as well—curious about how it decides which information qualifies as "highlights".
Cheers,
Xinran
-
P.S. I gave a presentation at Columbia Business School several days ago to a group of MBA candidates. I talked about how AI has been revolutionizing the prototyping process.
It was my first time back on the new campus since graduating from the School of Architecture over a decade ago. It made me feel old. Time flies.
As I watched the students hurrying from class to class, I thought of myself back then, idealistic, dreaming of making the world a better place with design. There's no place like school.
Interesting. Basically, basically it does transcription + post analysis to help manage user interview & generate research report?
Any good tool for offline use?