Pre-Seed Deal Sourcing with GitHub Data: A Practical Guide
How to use GitHub engineering signals to find pre-seed startups before they raise. Covers what pre-seed activity looks like on GitHub, signal patterns, and a step-by-step sourcing workflow.
Key Takeaway
Pre-seed startups are the hardest to find through traditional deal sourcing because they have no funding history, press coverage, or database entries. GitHub is one of the few places where pre-seed activity is visible: 1-3 contributors, rapid commit acceleration from a low base, and new repository creation. The key signal is disproportionate acceleration — a small team suddenly shipping at 3-5x their baseline rate, which typically indicates a product breakthrough or preparation for a first fundraise.
Pre-seed startups are invisible to most deal sourcing tools. They have no Crunchbase entry, no press coverage, no PitchBook profile. But many of them have GitHub activity.
Why GitHub Works for Pre-Seed Sourcing
GitHub is one of the few public data sources where pre-seed activity is visible. Before a startup has a pitch deck, before it has a website, the founders are writing code. That code — if the repositories are public — creates a data trail.
At VC Deal Flow Signal, we estimate startup stage from contributor count: pre-seed companies typically have 1-7 contributors. When we see a small team showing disproportionate acceleration — committing at 3-5x their baseline rate — something is happening worth investigating.
What Pre-Seed Signals Look Like
Pre-seed engineering activity has a distinctive pattern:
Low absolute velocity, high acceleration: A solo founder going from 5 commits/week to 25 commits/week shows +400% velocity change. In absolute terms, 25 commits is nothing. But the acceleration is the signal.
Infrastructure buildout from zero: New repositories appearing (3+ in 30 days) in a young organization suggests the founder is moving from prototype to structured development. This is classic pre-seed-to-seed transition behavior.
Contributor count jumping from 1 to 3-4: When a solo founder suddenly has co-contributors, they either found a co-founder, hired their first engineer, or attracted open source contributors. All three are positive signals.
A Pre-Seed Sourcing Workflow
- Check the sector rankings weekly and filter for startups showing "Pre-seed" stage estimation
- Look for companies with +200% or higher velocity change from a small base
- Open their GitHub organization — is the activity product-related?
- Check if the founder is active on Twitter, Hacker News, or Indie Hackers
- If signals align, reach out before anyone else knows the company exists
For the complete screening methodology, see the 7 engineering metrics every investor should track.