Reflections on research, design, and the tools that shape how we work.
I used to spend hours transcribing interviews. Now I spend those hours actually thinking about what people told me. That shift — from mechanical to meaningful — is what AI has genuinely given me as a researcher. But it's also not the whole story.
Read on LinkedIn →I used to keep my research notes in folders, spreadsheets, and sticky notes. Then I started using Dovetail — and the way I manage qualitative data changed completely. It surfaces patterns. It doesn't tell you why those patterns matter for your specific user, your specific product, your specific moment in the market.
Read on LinkedIn →I used to take notes during interviews. Real-time, in a notebook, while trying to listen. AI transcription tools changed that for me — not by replacing the interview, but by freeing me to actually be present in it. Transcription is not analysis. The real gain is the attention you get back during the conversation.
Read on LinkedIn →What if you could run a usability test with 100 participants in the time it usually takes to schedule 5? Maze largely delivers on that promise. But unmoderated testing tells you what people do — it rarely tells you why. Maze is most powerful when it's not used alone.
Read on LinkedIn →Most usability problems are visible. Information architecture problems are invisible — until they're not. Optimal Workshop helps you understand how people think about and navigate information through card sorting, tree testing, and first-click analysis. But it doesn't resolve the tension between user logic and organisational logic.
Read on LinkedIn →I've been using Miro for years. But in the last twelve months, it started thinking back. For service designers and researchers, this changes something fundamental about the workshop-to-insight pipeline. Miro AI is very good at organisation. It's not good at meaning.
Read on LinkedIn →These aren't niche research tools — they're general-purpose AI assistants I've woven into almost every stage of my research practice. From desk research to discussion guides, synthesis support to report drafting. Used well, they make me faster and sharper. Used uncritically, they flatten everything into something that sounds like research but isn't.
Read on LinkedIn →Seven tools. One recurring conclusion. AI is very good at the surface of research: the data, the patterns, the volume. It is not yet good at the depth: the meaning, the context, the consequence. The researchers who will thrive aren't the ones who resist AI or defer to it — they're the ones who know exactly when to use it, and when to trust themselves instead.
Read on LinkedIn →