Customer Listening, Encoded
"Engineering with Insight" is now a product: Lisnloop in beta
A few months ago I published Engineering with Insight, a field guide for engineers at early-stage SaaS companies who own more than just the code. The ones who talk to customers, frame problems, and make bets.
It resonated more than I expected. 🙏 And I started getting even more questions from engineers, like:
“Am I over-indexing on this one comment?” “What’s a better way to ask, ‘how much would you pay for this?’” “What should I do differently with my next call?”
So I’d write back with a take, or a reframe. Or we’d talk through the one question that would have changed the conversation. Sometimes I’d mark where they led the customer. I’d name the thing they missed.
But then…
I got very busy building these conversations into a product!
Introducing Lisnloop.
It’s the first product from Arbor — built from patterns across our client work. A simple chat interface, but underneath it's dense with product wisdom for product engineers working in early-stage SaaS. Opinionated, specific, narrow. All geared around customer “listening loops” (hence, the very available name/domain).
What it isn’t: Lisnloop doesn’t “do discovery” for you. It makes you do the work, so you build your gut, so you build better stuff. It doesn’t validate your idea so you can feel good.
What it is: A product sparring partner that makes you sharper at discerning customer signal. Lisnloop has your company and product context from day 1, thanks to a background Company Snapshot at signup, and then it uses your actual work from Linear to ground the conversations.
Let’s ask Lisnloop what it’s good at:
How engineers are using Lisnloop right now:
Before a customer call — Get specific about what you’re trying to learn and Lisnloop generates learning goals and a conversation outline. What assumption are you testing? What behavior would change your mind?
After a customer call — Paste the transcript. See where you led the customer, where you stayed surface-level, and Lisnloop builds a better outline for the next customer call. (Yes, it can pull customers insights from those transcripts, too. But if you're just harvesting quotes, you're missing the point. The goal is you getting better at asking the right questions.)
When you’re early on a project — Shrink the bet until it can teach you something this week. Connect Linear and start the chat from there; what’s the smallest thing you could ship that would build conviction?
When CS or Sales is feeding you “what they’re hearing” — Turn anecdotes into patterns you can actually build against.
When you have a pricing hunch but no signal — What are they paying now? What would they cut? Lisnloop pushes past “would you pay for this” (everyone says yes) to questions that reveal real willingness.
When you need internal buy-in — Frame the bet clearly enough that someone else can disagree with it. Shape the message before the meeting.
A note on how it’s built, since engineers care:
The stack is intentionally quite boring. Next.js, React, Postgres/Neon, Vercel. Commercial AI APIs with strict data handling — your conversations aren’t used for training. No hidden agents, no mystery flows. The whole thing is designed to be observable and auditable, because that’s table stakes if you’re asking people to think out loud.
More detail on the privacy page if you want to trace exactly how data moves.
(Ping me if you want to poke at the repo…or want to pitch in!)
Free for engineers. Capped to (hopefully) stay free.
If Engineering with Insight resonated, this is the next step. Bring a transcript from your last customer call, or a half-baked idea you’re not sure about yet.
Use it with your work, and watch your judgment change. You start to feel which problems are alive, where friction really is, and what “good enough to learn” looks like
If you’re an engineer working across the entire product lifecycle, try lisnloop.com →
🙌 Huge thanks to all the engineers who have pushed on Lisnloop over the past few months to help me shape it — both the code and the product (Igor and Matt, in particular!).
Things we’re looking at next: Other integrations/MCP servers (PostHog for data pulls to add quant data, CRMs for finding the right customers to chat with, etc), more interactive project spaces, pushing work to Linear, etc. So let me know what would be helpful!