Kickstarting Continuous Discovery as an IC PM
Here’s a playbook I’ve used to bootstrap continuous discovery with our customers. It’s transformed how our team builds products - and led to engineers joining calls, decisions that are grounded in real customer problems, identifying new user needs.
Here’s how to get started with as little overhead as possible (with huge credit to Teresa Torres for her work around continuous discovery).
First, why does continuous discovery work?
Continuous discovery helps you develop what I call “customer terroir” - that deep, contextual understanding that comes from regular exposure to your users’ environment. This goes beyond explicit feedback or feature requests; you start to internalize their workflow patterns, pain points, and unstated needs.
What You Need to Get Started
The barrier to entry is surprisingly low. Here’s your essential toolkit:
- CRM access (Iterable, Customer.io, or similar)
- A calendar scheduling tool (Calendly, e.g.)
- Video conferencing software (Zoom, Google Meet)
- Optional but recommended: AI note-taking tool like Granola.ai
Setting Up Your Research Pipeline
1. Create Your Email Outreach System
Set up a recurring email campaign in your CRM with these key elements:
- Target a small subset of customers weekly
- Implement a suppression rule to prevent repeated outreach (e.g., no contact in the last 3 months)
- Personalize the message and set reply-to as your email
Here’s a template that works well enough to get started:
Hey [Name],
I’m Ezra, a product manager at Dishwashers Co. and I work on the dishwasher you use. We’re always trying to improve our products, and I’d love to hear about your experience.
Up for chatting for 30 minutes? You can schedule time here: [Calendly link]
Best,
Ezra
2. Configure Your Calendar
Create a dedicated “User Research” meeting type:
- Limit scheduling to one session per week initially
- Link it directly to your video conferencing tool
- Pro tip: If you don’t have access to scheduling/shifting software, use a Google Form with time slot drop-downs, and close the form when you have the responses you need
3. Calibrate Volume
Expect something between a 1 and 10% response rate as a starting place. If you want a single conversation a week, email ~10 to 100 people. Generally I recommend sending to 10 people at a time for the first send, then increase gradually.
Make the Most of Each Conversation
This article is primarily about the playbook and workflow, so I’ll point you to further reading below for the most part. But a few things to keep in mind:
- Partner with designers or engineers when possible—share the load and increase coverage.
- Avoid asking leading or general questions. The tip from Teresa’s book that stuck with me was “Tell me about the last time you did ___.”
- Let people know you’ll be recording. (Don’t forget to record it.)
Get Summaries Down Quickly
Take the 15 minutes immediately after the call to write down what stood out. Look at the transcript and AI summary, bold your key takeaways.
Drop these notes, along with the recording link, into your team’s shared channel (do you have one for #customer-feedback? now is a good time to start one). Create that shared understanding.
Scale Yourself
Once you’ve got the basic rhythm down, you can automate even more of this:
- Use Zapier to route your recordings and transcripts to a Notion doc
- Auto-push summaries to your Slack channel
- Let AI help surface patterns across conversations
- Cut video clips of the key transcript moments
The goal isn’t to remove yourself from the process - it’s to spend more time on the conversations themselves and less on the logistics around them.
Further Reading
If you haven’t read it yet, Teresa Torres’s Continuous Discovery is a must read. As is Erika Hall’s “Just Enough Research.”
Wrapping Up
If you’ve implemented a system like this, I’d love to know how it went. Pro-tips? Things you’d change from this guide? Shoot me a note.
