Hello {{First name|Predictable Revenue community}},

vclist update: I've found a way to go from "tell me about your startup" to "find a list of investors actively investing in the space" to "show me who can make intros." It’s the happy path that just about every founder I interviewed wanted. If you’re raising and want to be a tester hit reply. Note - this time it will be a paid (not a huge amount) test, this thing burns tokens but it gets the job done.

Onto the newsletter…

No newsletter this week. Just a tool I built that I think you'll actually use.

I built a market research skill for Claude Code. You type one command, give it an industry and some competitor domains, and a few minutes later you get back a strategy document that would have taken you weeks to put together manually.

Here's the repo: https://github.com/collin128/market-research-skill/

The idea started with a tweet from Rimsha Bhardwaj (@heyrimsha) about using AI prompts for competitive research. Good tweet. My first thought was, "maybe I could turn that into a skill."

So that's what I did.

The skill fires off parallel research tasks using Exa's API, pulls in competitive intelligence, customer voice data, market dynamics, and moat analysis, then runs it through a five-stage analytical chain. You get competitive landscape, customer sentiment from Reddit and review sites, market trends, and a section called "The Unspoken Insight," which surfaces what actually separates the winners from the losers in a given market.

It also includes a Seven Powers analysis which is not very good yet but it gives you a general direction. I find it’s overly generous at applying the powers. This is a great tool for finding information about a market but do not outsource your thinking to an LLM. 

Total runtime: 2-5 minutes. Cost: roughly $1.50 to $5.50 per run.

I built this because I wanted to run it for RelyLoop, my startup solving email workflow problems property managers, but I didn’t want to sit and copy-paste a bunch of prompts into Claude. So I spent a few hours building and running the skill because I was too lazy to sit and spend 5 minutes copy-pasting.

I am aware my math doesn’t check out.

To be clear, this doesn't replace talking to actual customers or doing your own research. Nothing does. But it gives you dramatically better pre-work for customer discovery. You show up to your first conversations already knowing what the market looks like, what customers are complaining about, and where the incumbents might be weak. Better context means better questions.

If you use Claude Code, go grab it: https://github.com/collin128/market-research-skill/

You'll need an Exa.ai API key and Python 3. Setup takes about two minutes.

Run it against your own market. Run it against a market you're curious about. Run it against your competitor's market just to see what comes back.

In other news, I officially incorporated ReplyLoop this week and have been playing "front end guy" for the product alongside my co-founder Matthew. There's something deeply satisfying about looking at a UI, thinking "I hate that," and knowing I can fix it. That used to be reserved for engineers. The gap between technical and non-technical founders is shrinking faster than most people realize.

Back to regularly scheduled programming next week.

Collin

PS - shoutout to Paul for the glowing review of my book, The Terrifying Art of Finding Customers.