Free email course · 30 days · ~10 min/day
Four weeks. Seven GitHub signals in Week 1. Three real-pipeline candidates in Week 2. A 10-org watchlist + weekly Monday rhythm in Week 3. Alerts, MCP integration, and a custom-weighted composite in Week 4. End the month with an operational sourcing system that runs in ~25 minutes a week — drawn from the SSRN-published methodology covering 219 startup-period observations.
By Day 31 you’ll have a watchlist, a personal heuristic for your beat, an alerting layer, an MCP integration, and a 30-day retrospective on what the framework actually caught. No tool, no API, no warm intro required for the manual version.
This challenge works because you do the work, not because we email you well. Three pledges before you start:
Commit velocity
The simplest acceleration signal — 14-day vs 90-day ratio.
Contributor diversity
The bus-factor signal — Gini coefficient at month -3.
Dependents graph
The hidden GitHub page — cheapest external-adoption proxy.
README freshness
The cheapest leading indicator — a README rewrite is a pre-pitch.
New repo creation rate
The platform-buildout tell — SDKs, CLIs, demos shipping fast.
Issue-to-PR ratio
Engagement vs shipping — 1.5+ healthy, 0.7- firefighting.
The composite + the fast version
Six signals → one score → 4,200 orgs in 4 seconds.
Pick three real candidates
Three orgs from your own pipeline — one boring, one obvious, one wildcard.
Score candidate #1 end-to-end
The boring one — six signals, 15 minutes, one composite.
Score candidate #2 — the obvious one
The startup you already liked — does the framework agree?
Score candidate #3 — the wildcard
The unread one — where the framework earns its keep.
Compare all three — which signal carried the read?
Three composites side by side — and the signal you almost skipped.
Calibration — score a known recently-funded org
Pick a Series A from last quarter. What did the composite say at month -3?
Week 2 wrap — your first scorecard
Three composites, one calibration, one personal heuristic. That's a real artifact.
Build a 10-org watchlist
Ten orgs, one sector, scored once, queued for weekly recheck.
Set the weekly Monday rhythm
25 minutes every Monday, before email — the cadence that compounds.
Sector batch — 25 orgs in 25 minutes
When a sector heats up, sweep it broad before going deep.
Score-driven founder questions
Five questions, each seeded by a signal — sharper than 'tell me about traction.'
Score one portfolio company retroactively
Pick the worst-performing investment. What did the composite say at term sheet?
30-second pre-read for first meetings
One glance at the composite before you walk in — and you're already ahead.
Week 3 wrap — your operational system
Watchlist + Monday rhythm + sector-batch + question library + pre-read.
Set alerts on score moves
When does a score change matter? When commit velocity flips by ≥0.4.
The anti-signal — when 6/6 is wrong
Some orgs game every signal. Building your false-positive list is week 4 work.
Co-investor share template
Score → paragraph → ask. Three lines, one Slack message, one new co-investor convert.
MCP integration — running the framework from your editor
Six tools, one MCP server, zero context-switching out of your IDE.
Custom composite weights
Default is 1×6. Your beat probably needs 2× on one signal and 0.5× on another.
30-day retrospective — what the framework caught
Four weeks. Three candidates. One calibration. Ten watchlist orgs. What's measurable?
The Sunday digest as continuous practice
After Day 30, the rhythm continues — five orgs every Sunday, one composite per week.
Pre-graduation — which rung makes sense?
Free, €9.97/mo, or €1,997 — match the rung to the actual usage pattern from the last 30 days.
Graduation — the Stack Slide and one quiet decision
Thirty days. Six signals. One scorecard. One operational system. One decision.
Click any day to read it as a permalink — useful if you want to preview the curriculum before subscribing, or refer back after.
Comparable signal-walkthrough + applied-practice courses charge €69-€297 per lesson. Priced as paid lessons across the four phases, the 30 days break down to:
Week 1 — Learn (Days 1-7)
Each of the 7 atomic signals: commit velocity, contributor diversity, dependents graph, README freshness, new repo creation, issue-to-PR ratio, composite scoring.
Week 2 — Apply (Days 8-14)
Score 3 real candidates from your pipeline + 1 calibration backtest + your first scorecard artifact.
Week 3 — Synthesize (Days 15-21)
Build a 10-org watchlist, set the Monday rhythm, sector batching, score-driven founder Q&A, the 30-second pre-read.
Week 4 — Operationalize (Days 22-30)
Alerts, anti-signals, co-investor share template, MCP integration, custom weights, retrospective, graduation.
Free because the methodology is published (CC BY 4.0) at ssrn.com/abstract=6606558. The Challenge delivers the framework as a 30-day cadence and adds the operational practice around it — we charge for scale (the live engine across 4,200 orgs and the custom Sector Sweep), not for the framework.
/challenge/[slug] permalinks are the reference)Free. No card, no trial, no hidden upgrade pressure. Thirty emails are the course; Day 30 mentions three optional paths but you keep the framework either way.
~5-10 minutes a day. Week 1 is tight 5-min walkthroughs of one signal each. Weeks 2-4 stretch to 10-15 min/day as you score real candidates and build the watchlist + alerting + share template + MCP integration.
No. If you have ever opened a repo on github.com, you have everything you need. The exercises read what is already on screen and combine simple counts.
Panel analysis of 219 startup-period observations, published as GitHub Signals as Leading Indicators of Venture Fundraising on SSRN (abstract id 6606558). Each daily email maps to one signal in the paper.
The Sunday digest is the applied product — five named startups every week. The Challenge is the framework — the underlying signals so you can run the analysis yourself on any startup, not only the ones we surface.
Yes. Every day has a permalink — see the day-by-day list above, or jump straight to /challenge/commit-velocity. Subscribe to receive them on cadence, or read all seven straight through.
Methodology: /methodology · Paper: ssrn.com/abstract=6606558