A practical guide to using public GitHub data to evaluate startup engineering health, momentum, and fundraising likelihood.
After analyzing 219 startup fundraises, three GitHub signals stood out as predictors of upcoming funding rounds:
What it measures: The rate of code commits over time, specifically the trend (accelerating vs. decelerating).
How to read it:
- Accelerating (>20% QoQ growth): Team is shipping faster, likely hiring, product is gaining traction
- Stable: Healthy maintenance mode or steady growth
- Decelerating (>30% QoQ decline): Red flag. Could indicate founder burnout, technical debt, or team departure
Sector benchmarks (Q2 2026):
| Sector | Median weekly commits | Top quartile |
|---|---|---|
| AI/ML | 47 | 112 |
| Developer Tools | 89 | 203 |
| Fintech | 34 | 78 |
| Web3 | 56 | 134 |
| Healthcare | 22 | 51 |
What it measures: The number of active developers pushing code to the organization's repositories.
How to read it:
- Growing: Hiring is working, team is engaged
- Stable at >5: Established team
- Declining: People are leaving (check if key contributors stopped)
- Single contributor: Single point of failure risk
What it measures: Presence and activity of CI/CD, containers, monitoring, and IaC configs.
How to read it:
- Adding CI/CD, Docker, monitoring: Engineering culture is maturing, preparing for scale
- Removing configs: Troubling sign of either cost-cutting or team turnover
- No infrastructure configs at all: Early stage, higher risk
- Visit the startup's GitHub organization page
- Check repository activity (last commit date, contributor list)
- Review commit history for the last 90 days
- Look for
.github/workflows/,Dockerfile,docker-compose.yml - Check the Insights tab for contributor graphs and commit frequency
Tools like VC Deal Flow Signal automate this across 350+ startups simultaneously, scoring each on a composite engineering acceleration index and alerting when a startup's signals spike before a funding round.
The tool tracks 15 sectors including AI/ML, Developer Tools, Fintech, Web3, Healthcare, and more. Live data and sector benchmarks are available at the startup signal tracker.
- Bot commits: Some teams use bots that inflate commit counts. Always check contributor profiles.
- Monorepo bias: A monorepo with 20 services will have more commits than 20 microservice repos. Normalize by team size.
- Open-source vs proprietary: An open-source-first company will naturally have higher GitHub activity. Compare like with like.
- Private repos: Many startups do most work in private repos. GitHub activity is a floor, not a ceiling.
- GitLab/Bitbucket users: Some startups use alternative platforms. GitHub signals only cover GitHub users.
- Non-code work: Design, product, and ops work doesn't show in commits. Don't treat GitHub as the only signal.
The best approach combines GitHub signals with traditional diligence:
- Use GitHub velocity as an early indicator (47 days before round on average)
- Cross-reference with Crunchbase, LinkedIn, and product analytics
- Build a watchlist of accelerating startups in your thesis sectors
- Reach out when signals spike, before the round is announced
This guide is based on analysis published in an open SSRN paper covering 219 fundraises with pre-registered predictions and a transparent 0-for-10 first cohort.
CC BY 4.0
- VC Deal Flow Signal — Live startup momentum dashboard tracking 350+ startups across 15 sectors
- GitDealFlow — Deal flow intelligence tool for VCs and angels using GitHub signals
- Due Diligence Checklist — 52-point technical DD checklist for investors