VC List Part 3?

Hello Predictable Revenue community,

Book update: A huge thanks to those of you that were at the YC founder’s talk at Lightfield last week and brought up my book. The organizers were intrigued and reached out to invite me down. We’re working on something for January and I’ll have more details in the coming weeks. If that was you, drop me a line because I’d like to send you a thank you.

Note: If you’re raising or thinking of raising, check out the PS, I’m looking for a few folks to help me test something interesting. 

Onto the newsletter… 

The VC list

During the book launch, my goal was to gently bribe you all with something you thought was worth more than the $20 my book cost. I tried sharing my founder's sales playbook, GPTs that I use every day, and even guides that would help you prospect but none of those really moved the needle like the VC list. That list alone probably generated somewhere between 50-100 pre-orders. Nothing else came close.

Since building that list, I’ve helped a few of my clients tailor this list by creating a classifier prompt trained on their company and running it across the whole list. This knocks the 21k list of investors down to the few hundred most relevant firms. This has helped them get meetings with both top tier firms and niche firms they wouldn’t have found otherwise. 

I was relaying my experience to one of our board members and he asked what I’d charge to do this for one of his portfolio companies. I couldn’t give him a good answer. I like doing Clay work, but I don’t love repeating the same tasks over and over, even if they’re profitable. One of our sons is the same, if he wants to do something, he’s a total pro. If he doesn’t, he melts into a useless jelly blob. My board member encouraged me to rethink my position but I was deep in the book launch and didn’t have any spare cycles. 

After getting the Ticketmaster post off my chest, I found myself with a few spare hours on a Friday afternoon. I had a decision to make. Do I take the three hours and call it an early weekend or do I do something a little more fun? You’ll never believe it but I decided to sit down and noodle on this VC list idea. Just a prototype… something I could test with a few people.

Vectors! 

I hate list building, or at least, I used to. Back in the days where Zoom reigned (not the video call platform, yes I’m avoiding saying their full name) you had to pick from an often irrelevant set of filters to try and get access to the data that you’re looking for. It was a frustrating experience of memorizing SIC and NAICS codes and trying to find the ones most relevant to the group of companies you wanted. The problem was, once you found the perfect code, there would only be 15 companies that matched it. 

When Clay.com entered the picture, it gave us the ability to search by company size and geography, pull a huge list into a sheet, and then filter it with a classifier prompt. The prompt takes each company’s description, compares it to the type of company you’re targeting, and labels it as a good or bad fit. This makes finding good fits extremely easy. There are no secret prompts for me to share, just tell a LLM what you’re looking for and ask it to create a classifier prompt. 

Earlier this month, I helped a client who runs a vector database company build a list of investors, and they pointed out that my VC dataset would be a perfect fit for a vector database.  That comment had been sitting in the back of my mind ever since. Vector search is good at finding conceptual matches, not just keyword matches, similar to my Clay classifier system. Which is exactly what you need when you're trying to find VCs who would actually care about your company. I wanted a user to be able to paste in a short description of their company and get a list of the most relevant VCs and investors at each of the firms. 

Last Friday I sat down with Lovable, an AI coding tool that integrates tightly with Supabase, and built a pretty neat prototype in just a few hours. Supabase gives you Postgres with PGVector baked in, plus pg_cron for scheduling background jobs. Everything I needed in one place, without having to manage my own infrastructure. It was a much faster experience than trying to go from zero to one with Windsurf. 

I moved my Clay data into Supabase and started building. Three hours later, I had something running. You paste in a description of your company, it runs a vector search, and it returns a ranked list of VCs that match what you're building.

The first time it worked, I actually laughed out loud. It felt like magic.

Making it better

The problem with that first version was that it only searched against VC firm descriptions. Which works okay, but it's not great. It's still pretty surface-level.

I wanted to go deeper. So over the weekend, I scraped portfolio companies from investors' LinkedIn profiles - about 65,000 of them. Then I wrote a Clay agent to visit each company's website, figure out what they do, and generate a summary.

The reason for all this extra work is that company names by themselves don't mean much to a vector search. "Stripe" as a word doesn't tell you anything about payment processing. You need the actual meaning encoded, not just the name.

Now when you search, it's not just matching you to VCs who say they invest in your category. It's finding the specific investors at those firms who have backed companies similar to yours. If someone has a portfolio company that looks like what you're building and they're advertising that investment on LinkedIn, there's a decent chance they'll take your call. 

Technical headaches

Getting three different data types - VC firms, individual investors, and portfolio companies - all loaded into Supabase, embedded properly, and processing in the background has been harder than I expected. I keep running into edge cases with how Lovable handles larger datasets and background jobs.

But the core of it works. You can paste in your company description and get back a list of investors ranked by how well their past investments match what you're doing. Which is what I set out to build.

The pricing question

This is where I'm stuck. Most competitors charge around $100 per month just to access their database. Their datasets are smaller than mine - I haven't found anything with this combination of firm data, investor data, and portfolio company intelligence that isn’t thousands of dollars. 

Building the scraping infrastructure used to cost thousands or tens of thousands of dollars. Someone had to research each firm, track down the right contacts, figure out what they invest in. Now it's automated, which should make it cheaper.

My costs are low, hosting and embeddings don’t amount to much, and I want to charge a reasonable rate. But I also know most people won’t use this every day. They’ll sign up when they’re fundraising, run a bunch of searches, maybe export some lists, and then cancel until the next round. That’s exactly how I use a tool called PodEngine. They charge $100, which is fair for my use case, but I only need it two or three times a year, so I end up signing up and canceling repeatedly, which I’m sure is frustrating for the founders.

I’d actually prefer a slightly higher price, say $300, that gives me annual access, even if the number of exports is limited.

So what should I charge? A hundred dollars per month or three hundred for a year? Fifty to undercut everyone else? Twenty-five with an annual plan?

I genuinely don't know. And I'm asking because I'd rather hear what feels fair than guess wrong and either leave money on the table or price myself out of the market.

One other idea I've been thinking about: offering a free month to anyone who buys the book and posts a review on Amazon or LinkedIn. Is that a reasonable exchange, or does it come across as transactional in a way that feels off?

The bigger picture

A while back, a friend walked me through how he structured his business. Free webinar at the top, then a twenty dollar ebook, then a two hundred dollar course, then a thousand dollar event, and finally a five thousand dollar monthly service. Each step naturally led to the next. Their marketing team could focus their efforts on driving folks to the top of the funnel and then slowly work their way down. They had their math and process super dialled and the results showed. 

Predictable Revenue doesn't have that. We have a twenty-dollar book and then expensive consulting. There's no middle ground. Nothing between "read this" and "hire us."

I tried creating something with the monthly Q&A calls for founders, but I could never bring myself to charge for them. I enjoyed helping folks and didn’t want money to get in the way. This tool might actually work as that middle step. It provides value without requiring my time, and it creates a clearer path from someone reading my content to becoming a client.

And selfishly, it might drive a few more book reviews.

If you have thoughts on any of this - the pricing, the positioning, whether this even makes sense as a product - I'd genuinely appreciate hearing them. Just hit reply.

Collin

PS - I need a few people to test this. I'm not ready to launch this broadly. There are still rough edges I need to smooth out. But I'd like to get it in front of three or four people who are actively raising and are willing to give me feedback. If that sounds like you, hit me back. 

PPS - we’re up to four reviews, big thanks to Sharon for sharing her thoughts. Grab your copy of The Terrifying Art of Finding Customers at Amazon, Indigo, and Barnes and Noble.