Gingiris commited on
Commit
d9a1b20
·
verified ·
1 Parent(s): 1dfd945

add README star loop and geography benchmarks from founder interviews

Browse files
Files changed (1) hide show
  1. README.md +41 -0
README.md CHANGED
@@ -152,6 +152,47 @@ clawhub install gingiris-github-star-growth
152
 
153
  ---
154
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
155
  ## English
156
 
157
  ### What's Inside
 
152
 
153
  ---
154
 
155
+ ## 🆕 What's New (July 2026): The README Star Loop and Geography Benchmarks
156
+
157
+ ### The README Star CTA Loop
158
+
159
+ Most projects leak stars at every exit. The fix is a closed loop: map every exit path a visitor can take — website, docs, community — and link each one back to the GitHub README, with a star CTA sitting at every backlink. Website footer points to the README. Docs pages point to the README. Discord welcome message points to the README. No matter where traffic enters or wanders, every route passes the star prompt before it leaves.
160
+
161
+ The same discipline applies upstream. Route every channel — Reddit, Twitter, Medium, directories — to the GitHub repo link, **not** the website. The repo page is where the star button lives; the website is where stars go to die. During launch week, check GitHub Insights (Traffic → Referring sites) daily and double down on the channels that actually send traffic instead of the ones that merely feel active.
162
+
163
+ ### A Verified, Never-Bought Trajectory
164
+
165
+ What does a fully organic star curve look like? AFFiNE's (self-reported):
166
+
167
+ | Milestone | Timeline |
168
+ |:----------|:---------|
169
+ | ~1,000 stars | First 72 hours |
170
+ | ~6,000 stars (80% overseas) | Week one |
171
+ | 10,000 stars | 43 days |
172
+ | 60,000+ stars | 24 months |
173
+
174
+ Not a single star was purchased. Multiple investors ran their own crawlers over the stargazer list during due diligence to verify the growth was organic — and it held up. Bought stars are not just useless; they are a landmine that detonates exactly when the numbers matter most.
175
+
176
+ ### Geographic Health Benchmarks
177
+
178
+ From studying the stargazer geography of VSCode, Vue, and AppFlowy, a healthy global open source project looks roughly like:
179
+
180
+ | Region | Healthy Share |
181
+ |:-------|:--------------|
182
+ | United States | ~19-21% |
183
+ | China | ~19-21% |
184
+ | Top EU countries (combined) | ~10-15% |
185
+
186
+ First-week discipline is what protects this ratio: English-only communication, and no sharing to domestic WeChat Moments. The cautionary tale — once AFFiNE's China share spiked to ~35% in week two, the ratio never came back down. Geography mix is path-dependent; you only get one clean launch window.
187
+
188
+ ### Reddit Funnel Benchmark
189
+
190
+ For calibration: tens of thousands of subreddit visits converted into 2,000+ stars. If your Reddit-to-star ratio is far below that, the problem is usually the landing link (website instead of repo) or a missing star prompt at the destination.
191
+
192
+ *Distilled from 12 founder podcast interviews (2024-2026), figures self-reported.*
193
+
194
+ ---
195
+
196
  ## English
197
 
198
  ### What's Inside