The one where I finally found a real productivity breakthrough
Use case where AI really helps
At last - a real productivity gain has arrived. Not theoretical, not “AI-for-the-sake-of-AI,” but an actual workflow improvement where AI is genuinely excellent: consolidating user feedback from interviews and turning it into crisp insights.
If you’re a PM, you know the drill. We run user interviews often, and a lot - validating willingness to pay, stress-testing a new feature idea, or collecting product feedback. Before AI, the process looked something like this:
Recruit users
Prepare a script
Run 5–10 interviews
Review raw notes and consolidate insights
For me, steps 1 and 4 were always the most time-consuming. And recently, AI saved me hours on step 4.
Why not to use a dedicated AI tool for user interview?
There are plenty of specialized AI tools out there, to name a few:
Fireflies AI - AI powered note taker
Dovetail - AI platform that unifies multiple sources of feedback (interviews, support, etc.) into one database of insights
Looppanel - AI powered note taker, transcriber and analyser.
But here’s the catch: I don’t want to use a dedicated tool for every aspect of my work. I’d end up managing a small SaaS zoo while doing my PM work. No, thank you.
My rule: simplify the stack, not complicate it. So here’s the exact algorithm I used for user interviews. Plain and simple.
AI-Powered interview with no tools!
1. Define the target audience
Manual thinking required. No AI magic here.
2. Recruit users
I’m in B2B, where AI can’t really help with recruitment yet. Still manual.
3. Run the interview and record it
This is where things get interesting. I used the standard iPhone Voice Memos app to record the interview. Nothing fancy, no special tools, no subscription, no tokens etc. Voice Memos now has built-in speech to text transcription. It’s free, it’s fast, and it’s surprisingly accurate.
4. Generate summaries with AI
I uploaded the transcript + my interview script (for context) into our company-approved AI tool (you can use chatGPT, Claude, etc.). It produced:
a clean summary for each interview
top three cross-interview insights
5. Review and refine
A bit of manual work, but now I’m reviewing structured insights, not wrestling with raw notes.
So… was it worth it?
My rating: 5/5.
Actual time saved: 5–10% per week.
If you work on a single product, your savings may be smaller as you don’t run interviews every week. But I work across multiple products, so interviews are basically a weekly ritual. And this workflow now feels like a big helper.
Next time, I’ll tell you whether AI agents help too. Stay tuned!