After more than a decade as a user researcher, I’ve never found a digital analysis tool that fits the way I work. So I’m building one. Stitchwork is a qualitative analysis tool for researchers who need to interpret user interviews, usability tests and field studies quickly, without cutting corners. By selectively using features of modern AI, it allows you to work faster, while avoiding the downsides of having AI do all your thinking for you.
Why I’m building this
Doing things the old way, with pens, paper, highlighters and sticky notes, still works. It’s how researchers in academia and industry have done it for decades, so you can draw upon their tried-and-tested methods. But applying these methods rigorously is mentally demanding and slow going. They were developed by researchers who could spend weeks or months on analysis. When you have time, this can be an incredibly satisfying process. But in modern product teams you might have a couple of days before design and development move ahead, with or without your research.
Digital tools haven’t solved that problem for me. Some replicate the offline process with virtual sticky notes and highlighters. This isn’t much faster than the real thing, although it’s great for remote teams. Other tools, aimed at academia, have steep learning curves that assume a deep background in research theory. And then there are tools that are easy to pick up and quick to use for analysis, but strip away too much of the structure and rigour. They only provide a shallow descriptive analysis of the data instead of a deep interpretation. This fast, shallow approach has become more widespread with the developments in AI since late 2022. Many tool providers have simply layered AI on top of their existing products. They’ve removed the human element from analysis and turned researchers into data collectors for the machine. That brings a whole new set of problems, from hallucinations and ethical concerns, to analysis that lacks the experience, expertise and empathy of a human researcher. These are topics for future posts, but point to why a different approach is needed.
What Stitchwork is
For all the hype and pitfalls surrounding AI, the underlying technology is powerful. It’s unprecedented in its ability to rapidly process huge amounts of text, video and audio, and recognise patterns across this data. If properly managed, this can provide a massive boost to the speed and scale at which we exercise our analytical skills.
Stitchwork helps you quickly turn raw data into rigorous insight. It takes that underlying data-processing ability of AI and leverages it to handle the mechanical parts of analysis. You import your data, and Stitchwork helps you explore it. It surfaces recurring ideas and patterns in the data, freeing up your time to interpret their meaning and relationships. Unlike general AI tools, it draws on established qualitative methods from industry and academia, so your analysis stays deep and reliable. At every step, you decide which concepts, themes and findings are valid, and define what they mean. Everything stays linked back to the original quotes so you can trace every insight to its source.
This means:
- No more second-guessing whether you analysed the data “the right way”. Stitchwork walks you through a rigorous analytical process.
- Fewer debates over research holding up delivery.
- No more late-night analysis before a deadline because you ran out of time.
- Research team of one? Work through larger datasets as if you had a few extra pairs of hands.
- Research team of many? Keep everyone aligned within an analytical framework while drawing on their different perspectives.
- Spend your time on the human parts of research: sense-making, storytelling and persuasion.
- Get more insight from the data you already collect. You’re often just scratching the surface; now you can go deeper.
What’s happening next?
The seed of this idea was my own frustration with current user research analysis tools. But I know from years of training, managing and working alongside user researchers that plenty of people share this frustration. If that sounds like you, then what I’m building might be of some help.
Stitchwork is still in early development, and many of the specifics are being worked out. I’ll share updates on this blog as things progress. If you’d like to be among the first to know when Stitchwork launches, you can sign up for the waiting list. Or, if you have any thoughts you’d like to share, feel free to drop me an email.