How to Find Pre-Seed Startups to Invest In: 9 Sources Ranked by Signal Quality

"How do VCs find startups to invest in?" has a boring true answer and an interesting one. The boring answer is intros, events and inbound. The interesting answer is that a growing share of early investors read public signals systematically, before a founder has a deck.

This guide ranks nine sources by how much each one tells you, shows how to combine them, and gives you a weekly routine. It sits alongside our more general piece, How to Find Startups Before They Raise, and goes deeper on ranking.

The Principle: Score the Evidence, Not the Buzz

A pre-seed company has no revenue history and often no funding record. So you are judging a handful of public events. Two rules keep that honest.

Rule 1: independence. Signals from different kinds of sources are worth more together than the same kind repeated. A company listed on six launch platforms in one month has one kind of evidence, repeated. A company with an accelerator selection, a launch and a first hire has three.

Rule 2: freshness. Evidence should fade. AlphaScout halves the weight of each signal every 30 days and caps how much any one source can contribute, so a single noisy source cannot carry a company.

The Nine Sources, Ranked

Rankings depend on your thesis, but across most early-stage venture work this order holds up.

1. Verified Revenue and Revenue Growth

The strongest pre-seed signal is customers paying. AlphaScout gives revenue growth and verified revenue some of the highest point values in its table (+16 and +14 respectively), and adds bonuses when revenue grows 30 percent a month or more. It is also the rarest signal and the hardest to see from public data, which is why partner revenue data matters when it is available.

2. Accelerator and Fund Portfolio Selection

A program chose these founders. That is a third party doing a first screen for you. It is also the signal most likely to be shared with other investors, which is why AlphaScout caps companies already backed by an accelerator or fund, and why you may prefer to meet them earlier in the batch. See Accelerator Batches as a Deal-Sourcing Channel.

3. Product Launches

A launch on a launch platform, a Hacker News "Show HN" post or a product directory shows that something has shipped. It earns the largest single point value among events (+18 in AlphaScout's rules). The risk is volume: launches are cheap, and many are side projects. Treat a lone launch as a "watch," not a "meet."

4. Developer Traction

A new open-source repository, a model or an MCP server that gains users quickly is a strong signal for developer-facing companies. Look at velocity, not totals: GitHub stars up 50 percent or more in two weeks is flagged as fast-growing. Read Open-Source Traction as an Investment Signal for the vanity metrics to ignore.

5. First Hires

A first job post, especially for engineering or sales, means the founders expect to grow. It is worth +10 points in AlphaScout's evidence table, with another +4 for hiring as an early-stage adjustment. Hiring fifty people at once is the opposite signal, and it caps a score because that company is not pre-seed.

6. Grants and Crowdfunding

Non-dilutive grants such as SBIR and EIC awards (+10, with a further +6 for a non-dilutive grant) show that expert reviewers backed the work. Companies raising from the public through crowdfunding offerings (+10) are signaling that they need capital and are willing to disclose it.

7. Private-Placement Filings

In the United States, a company that has sold securities under a Regulation D exemption files a Form D within 15 calendar days after its first sale. It can be the first public sign a round has started. It is a legal notice, not a press release, and many financings never produce one. See What SEC Form D Filings Tell Investors.

8. Company Registrations and Domains

A new registration or a new domain says a company exists. Alone it says nothing about quality, but combined with other signals it dates a company precisely. AlphaScout also checks how long a website has been online: anything online for three years or more is capped, because it is not new.

9. Press and Directories

Press mentions and directory listings are the weakest signals. A directory listing earns just +3 points, and a company that appears only on self-submitted directories is capped. Heavy press coverage is a penalty (-8 for three or more articles), because it means the company is already known.

What Takes a Company Out

Ranking is also subtraction. AlphaScout's rules lower or cap a score when:

• The company has already raised $5 million or more (-12), or $20 million or more (-35)
• It has no website (-10), or the site is down or parked
• It is already backed by an accelerator or fund
• It has no new launch or traction yet
• The AI check says it is not a venture-style startup, such as an e-commerce store

These are the same filters a good analyst applies in their head. Writing them down makes them consistent.

Why Pre-Seed Sourcing Triggers Fear and Greed

Pre-seed investing sits at the intersection of two powerful emotions.

Greed, the pull of being first. Early companies are cheaper to enter and the rewards for being right are larger. That pull can make you chase the loudest launch of the week.

Fear, the dread of regret. Regret aversion is the tendency to avoid decisions we could later blame ourselves for. In sourcing, it shows up as the pit in your stomach when a company you passed on, or never saw, announces a round.

Both emotions push you toward bad process: chasing noise, or anchoring on whichever tool everyone else uses. A written scoring method and a fixed weekly review are the antidote. They turn a mood into a procedure.

The 30-Minute Weekly Routine

1. Collect (10 minutes). Open your feed or sources and scan only what is new this week.
2. Merge (5 minutes). Group mentions of the same company. Anything that appears in more than one kind of source goes to the top.
3. Screen (10 minutes). Apply your thesis. Drop anything out of stage, sector or geography immediately.
4. Decide (5 minutes). For each remaining company: reach out, watch or pass, and write one line on why.

That is thirty minutes if software has already done the collection and merging. Doing it manually takes several hours a week, according to our own estimate, which is exactly why most investors skip it.

Doing It With AlphaScout

AlphaScout runs steps one and two continuously. It reads 200+ public sources in 50 countries, ties every signal to one company by its own domain and strong identifiers such as an accelerator slug or an SEC CIK, and ranks the result with published rules. A company needs a gem score of 25 or more and a working website to qualify. For the week of 28 September to 4 October 2026, 48 of 10,696 new companies had qualified at the time of writing: under half a percent.

Every score shows its working: which signals moved it, from which source and how fresh they are. The scoring section shows the full table, and the free sample shows a real week of output. The live feed is $499 a month or $4,990 a year, with teams paying per seat.

Frequently Asked Questions

How Do VCs Find Startups to Invest In?

Most rely on referrals, events and inbound. Increasingly, investors add systematic monitoring of public signals such as launches, repositories, accelerator batches, hiring and filings.

How Do I Find Pre-Seed Startups as an Angel?

Start with free public sources: launch platforms, GitHub, accelerator pages and regional press. A weekly routine matters more than any single tool. See Building a Repeatable Deal-Sourcing Process for a Small Fund.

What Is the Best Startup Discovery Platform?

It depends on what you want to see first. Founder detection tools, enterprise graphs and evidence-ranked feeds all take different approaches. Our alternatives guide compares them.

Is One Signal Enough to Reach Out?

Usually not. Treat a single signal as a reason to watch. Two or more independent kinds within a month justify a closer look.

See the Method on Real Data

You can test every claim in this article in ten minutes. Open last week's hidden gems, check the evidence on a few companies, and judge the list yourself. No account is needed.

This article is for information only and is not investment advice. Signal values quoted are from AlphaScout's published scoring rules as of October 2026 and may change.