In partnership with
👋 Hey Friends!
You ran the discovery call.
You asked good questions. The customer said interesting things. You took some notes.
Then you went to your next meeting, and by Thursday you couldn't remember exactly what they said or why it felt important.
Sound familiar?
This isn't a listening problem. It's a system problem.
Most product teams treat discovery as an event. A call you schedule, run, and close. The insight is supposed to live in your memory or a doc somewhere until it becomes relevant.
It almost never does.
Why the call isn't the hard part
Here's something I've seen consistently: the teams who build the best products aren't the ones who run the most customer calls. They're the ones who have a system for what happens after.
The call surfaces the signal. The system is what turns it into something your team can actually act on.
Without the system, you're collecting data with no way to find the pattern. And patterns are everything. One customer frustrated by onboarding is an anecdote. Six customers describing the same friction point in different words is a roadmap priority.
The difference between those two things isn't research budget or headcount. It's process.
💡 The reframe: Customer-led product management isn't about running more calls. It's about closing the loop between what customers say and what your team builds.Most teams are good at the first half. Almost no one is consistent about the second.
Between meetings, speak your follow-ups. Done before the next one starts.
You have seven minutes between calls. That's enough time to type one email or dictate five.
Wispr Flow turns your voice into clean, professional text inside any app. Walk out of a meeting, speak your action items, follow-ups, and notes — Flow formats everything and you paste it where it needs to go. Email, Slack, Notion, your CRM.
Works on Mac, Windows, and iPhone. 89% of messages sent with zero edits. Used by teams at OpenAI, Vercel, and Clay.
The system that actually works
Here's the workflow I've landed on after building this from scratch more than once. For each step, I've noted what you can hand off to automation and what still needs your judgment.
1. Record everything, note-take nothing
Automate: Recording and transcription. Tools like Fathom, Fireflies, or Otter handle this automatically once set up. Zero effort on the call.
Do manually: Nothing. This step is fully automated. Your only job on the call is to be present.
This sounds small. It changes everything about how you show up.
2. Run a weekly transcript review
Automate: Drop your transcripts into Claude, ChatGPT, or NotebookLM and prompt it to surface recurring themes, friction points, and notable quotes. Takes two minutes.
Do manually: Decide what's actually signal vs. noise. AI will surface patterns but it doesn't know your roadmap context, your current bets, or what your team has already ruled out. That filter is yours.
3. Build a simple feedback taxonomy
Automate: Once your buckets exist, you can prompt AI to tag each theme to the right category as part of your weekly review.
Do manually: Creating the taxonomy itself, and updating it when your product evolves. This is a judgment call about what actually matters to your team — no tool can make it for you.
4. Connect themes to decisions
Automate: Nothing. This is the human layer and it has to stay that way.
Do manually: Looking at what's accumulated, deciding what rises to a priority, and making the case to your team. The AI gave you the pattern. You decide what it means.
This is the step most teams skip. They collect, review once, and never return to it. The value is in the accumulation and the judgment you bring to it every month.
🎯 Try this: After your next customer call, spend 15 minutes with the transcript and write one sentence: "The real issue this customer described was..."
Not what they said. What they meant.
Do that for three calls. See if the sentence starts to repeat itself.
Accio Work: the AI Agent team that runs your business
Meet Accio Work—the agentic workspace for business owners and solopreneurs. Our smart agents handle sourcing, supplier negotiation, store management, and marketing on autopilot. Powered by Alibaba.com data, we turn ideas into action instantly. No setup, no hassle—just seamless execution while you stay in control and focus on growing your business.
The bigger idea
There's a version of PM where you build what you think customers need, ship it, then check if it worked.
And there's a version where the customer's experience is the first input, not the last check.
The second version isn't harder. It's just more deliberate about what gets attention before the sprint starts.
Every team I've seen close a meaningful retention gap has done it by finding something specific that was quietly frustrating people. Not through a big research project. Through consistent, low-effort listening that actually got reviewed.
The call is just the beginning of that loop. The system is what closes it.
- Stef
P.S. If you're an early PM trying to build this system without a research team or big budget - this is exactly something we can work through in 1:1 session. Book here.