Analytics

Every brand gets a badge analytics view in the dashboard: how many times its badges rendered, which subjects (repos, packages) drive the most, where they render, and a 30-day trend. Analytics are a Pro feature.

Open it from Dashboard → Brands → Analytics on any brand.

What you see

• Total renders (last 30 days) — a floor, not a total (see the caveat below).
• By source — where the badge was rendered: github-camo, npm, docs, or direct.
• Top subjects — the repos, packages, or other subjects your branded badges point at, ranked by render count.
• 30-day trend — a per-day bar chart of render volume.

How it's measured

A badge render is counted server-side, fire-and-forget, on the same path that serves the image — it never adds latency to badge delivery. Each event carries the resolving brand, the badge provider, the subject, and a derived source. Counts roll up per day into a table your dashboard queries directly.

Only badges that resolve a brand (via ?brand= or /b/{slug}/...) contribute to that brand's analytics.

The Camo undercount caveat

GitHub proxies every README image through its Camo cache. When many viewers load the same README, Camo often serves a cached copy without hitting shieldcn, so those views aren't counted. Your real render count is therefore at least the number shown — never less, often more. The dashboard labels totals as "at least N" for this reason.

Privacy

Analytics store no personally identifiable information. Referers are reduced to a host-level source bucket, and no IP addresses or visitor identifiers are retained.