Data sources & provenance
Every size recommendation on ShoeSizeguide.com carries provenance: where the number came from, when it was last verified, and how confident we are. We do not copy competitor charts — we triangulate cited sources and show disagreements honestly.
Source types
- official_chart
- Published brand size charts (Nike.com, Adidas.com, etc.). We anchor on foot length in mm, not reproduced logo artwork.
- lab_measure
- Third-party lab measurements (e.g. RunRepeat, Solereview) cited with attribution and URL — never copied wholesale.
- review_aggregate
- Aggregated buyer fit reports (runs small / true / large) with sample size N, sample window, model/brand scope and collection notes when available. We use these as directional evidence, not as direct measurements.
- community_feedback
- Anonymous first-party «how did it fit?» submissions from our fit-feedback form. Highest-trust source once volume grows.
Review aggregate rules
- Aggregates must describe the collection method, sample window and scope when known.
- Aggregates can support confidence, but they do not override lab data or brand-published charts by themselves.
- When aggregate signals disagree with chart/lab sources, confidence is reduced and the disagreement remains visible.
- We do not copy review text; we store only derived fit direction, sample size and provenance notes.
What we do not do
- Claim a single «exact» size exists at brand level (physically impossible — US 9 spans 255–281 mm).
- Reproduce trademarked brand charts or logos in embeddable widgets.
- Index model pages without eligible data (confidence ≥ 0.6 and provenance).
- Hide source disagreements — when charts conflict, we show the range.
Freshness
Each record includes captured_date (ISO date of last verification). We review
top-brand records monthly, re-check broader official chart sources at least quarterly, and
keep the visible per-record date as the source of truth for freshness.