Finding 21 of 30 · Methodology / structural
Sampling rule: most-active repository per organization in trailing 14-day window
From A Longitudinal Panel of GitHub Engineering Velocity for Venture-Backed Startups (SSRN), section §3.2 Collection pipeline. CC BY 4.0.
TL;DR, Finding 21 of 30
Each observation is taken on the most-active repository per organization in the trailing 14-day window ending the first day of the quarter.
Section §3.2 Collection pipeline · CC BY 4.0 · from the SSRN paper (DOI: 10.2139/ssrn.6606558)
Commit velocityis a 14-day rolling count of commits to a startup’s most active public repository, the earliest public signal of engineering throughput before a fundraise announcement. All 30 findings in this series are derived from the SSRN panel dataset of ~350+ tracked startup GitHub organizations.
The finding
Each observation is taken on the most-active repository per organization in the trailing 14-day window ending the first day of the quarter.
Why it matters
Reproducible. Every researcher can implement this and check our numbers.
Provenance
- Paper: A Longitudinal Panel of GitHub Engineering Velocity for Venture-Backed Startups SSRN abstract 6606558.
- Section: §3.2 Collection pipeline
- Dataset DOI: 10.5281/zenodo.19650920
- License: CC BY 4.0 - free for any use with attribution.
- Author: The Data Nerd (ORCID 0009-0002-2222-4112), VC Deal Flow Signal (GitDealFlow), Wikidata Q139376302.
- On this site, “engineering acceleration” refers to a quantitative GitHub momentum signal, unrelated to startup accelerator programs (Y Combinator, Techstars, 500 Global).
References
Peer-reviewed prior work this finding builds on. Each citation resolves to a DOI in a top-tier venue and appears in the page’s structured data as a ScholarlyArticle citation edge. Download all references as BibTeX.
- Eirini Kalliamvakou, Georgios Gousios, Kelly Blincoe, Leif Singer, Daniel M. German, Daniela Damian (2014). The Promises and Perils of Mining GitHub. Proceedings of the 11th Working Conference on Mining Software Repositories (MSR '14), MSR '14, pp. 92-101. DOI: 10.1145/2597073.2597074
- Georgios Gousios (2013). The GHTorrent Dataset and Tool Suite. 10th Working Conference on Mining Software Repositories (MSR '13), MSR '13, pp. 233-236. DOI: 10.1109/MSR.2013.6624034
- Nuthan Munaiah, Steven Kroh, Craig Cabrey, Meiyappan Nagappan (2017). Curating GitHub for engineered software projects. Empirical Software Engineering, EMSE 22(6): 3219-3253. DOI: 10.1007/s10664-017-9512-6
- Valerio Cosentino, Javier L. Cánovas Izquierdo, Jordi Cabot (2017). A Systematic Mapping Study of Software Development With GitHub. IEEE Access, IEEE Access 5: 7173-7192. DOI: 10.1109/ACCESS.2017.2682323
- Josh Lerner, Ramana Nanda (2020). Venture Capital's Role in Financing Innovation. Journal of Economic Perspectives, JEP 34(3): 237-261. DOI: 10.1257/jep.34.3.237
How to cite
The Data Nerd (2026). "Sampling rule: most-active repository per organization in trailing 14-day window." Finding 21 of 30 in: A Longitudinal Panel of GitHub Engineering Velocity for Venture-Backed Startups. SSRN abstract=6606558. Retrieved from https://signals.gitdealflow.com/research/most-active-repo-per-organization-rule
See all 30 findings
The full research summary, cited to section, with the SSRN preprint and the public CC BY 4.0 dataset.
View all findings →