How we build this — and what we won't.
Anonymization. No single practice's data is identifiable in any benchmark. We require n≥30 campaigns per segment before publishing a number. Outlier trimming at 5th–95th percentile before median computation.
No cross-tenant peek. usdpr. practices opt in individually to the benchmarks program. A practice that opts out has its data excluded from every aggregate — not just the public dashboard. Data commitments.
License. CC-BY-4.0. Cite with attribution: "usdpr. Public Benchmarks (2026-Q1)". CSV download link coming 2026-Q3.
Next update. 2026-Q3 · Q2 2026 data. Announced on changelog, press kit, and in the quarterly transparency report.