What is Developer Signal Score? A 2026 Definition + Examples
Developer Signal Score — A Developer Signal Score is a numeric rating, typically 0–100, that measures how strongly an individual engineer's recent GitHub activity indicates purchase intent for a specific technology category. It combines event volume, event type, recency, and topical relevance into a single comparable number. Where a lead-scoring model in a CRM reasons about firmographics and form fills, a Developer Signal Score reasons about code — pull requests, issues, commits, stars, and forks on repositories relevant to the category you sell in. A score above 70 typically flags a developer worth personal outreach within 48 hours.
Quick definition
- Numeric rating (usually 0–100) for a single developer's recent GitHub activity
- Combines volume, event type weighting, recency decay, and relevance
- Higher weight on pull requests and issues than on stars and watches
- Time-decay applied so a star from 60 days ago counts less than a PR today
- Topic-scoped: the same developer can score high for one category and zero for another
- Works per-developer, not per-account — an individual engineer is the unit
- Typically refreshed every 15 minutes against near-real-time GitHub event data
- Used to rank outbound lists, trigger alerts, and gate automated outreach
How developer signal score works
The score starts by counting GitHub events the developer generated in a rolling window — typically the last 30 days. Each event type carries a different weight. A pull request might count 10 points; an issue 8; a fork 5; a commit 4; a star 1. Stars alone are weak; active contribution is strong.
A recency multiplier is then applied. Events within the last 7 days are doubled; events in days 7–30 count at 1x; older events decay toward zero. The idea is simple — a buying signal is hottest the moment it happens, and a three-week-old star is closer to noise than a PR submitted this morning.
A topical filter scopes the count to a category. LeadCognition ties each monitored repository to a technology tag (for example, "observability", "feature-flags", "vector-databases") so the same developer can show a score of 84 against your category and 12 against an adjacent one. This prevents hot activity on an unrelated repo from polluting your ranked list.
Finally, a normalization step rescales the weighted sum to 0–100 against a live distribution of other developers in the same window. A score of 70 means the developer is in the top quartile of recent activity for the category you're monitoring, not a fixed event count.
Examples
Example 1 — DevTool outbound. A company selling a CI/CD product monitors five competitor repos plus its own. Developers who open a PR on a competitor and then fork your repo score near 95. The SDR team receives them as a ranked Slack digest every morning.
Example 2 — PLG expansion. An API-first database company scores every developer who opens an issue asking about SSO, enterprise pricing, or audit logging on its open-source repo. Scores above 80 trigger an automatic alert to the account owner so outreach goes out within the hour.
Example 3 — Recruiting. A technical recruiter scores contributors to Kubernetes operator projects. Developers with a score above 70 have merged multiple PRs in the last 30 days, demonstrating active skills far beyond a resume.
Related concepts
Related glossary entries
Further reading
- Developer signal intelligence — the complete guide
- Why GitHub activity is the strongest buying signal
- Stars vs. PRs vs. commits — which signal matters
Related tools
FAQ
What range is a Developer Signal Score?
Most implementations use a 0–100 scale, with 70+ flagged as high-priority outreach and 40–70 flagged as warm. The scale is normalized against the live distribution of developer activity in the category you monitor.
How is a Developer Signal Score different from a lead score?
Traditional lead scores reason about firmographics and form fills. A Developer Signal Score reasons about code — pull requests, issues, commits, and forks on repositories in a specific technology category. It is available at the individual developer level, not just account level.
How often does a Developer Signal Score refresh?
LeadCognition refreshes scores every 15 minutes against GitHub event data. Because recency carries the largest weight, even hourly refresh intervals measurably lower conversion rates.
Can a developer have multiple scores?
Yes. The score is always scoped to a technology category. The same developer can hold a score of 84 for "observability" and 12 for "feature flags" depending on which repos they engage with.
See also
Browse the full LeadCognition glossary or visit the 36-answer FAQ for site-wide coverage. If you are specifically evaluating tools, start with the free tools or the sales-tool comparisons.