Methodology
How Duyuz measures Digital Index and benchmark samples — fetch rules, limits and what we refuse to invent.
How Duyuz measures Digital Index and benchmark samples — fetch rules, limits and what we refuse to invent.
Duyuz sample, not a census.
Methodology
Every Index or benchmark on duyuz.com starts from a disclosed list and one public homepage GET. We count successful fetches as n, list failed attempts separately and publish JSON. No browser rendering, no JavaScript execution, no invented Core Web Vitals. PageSpeed lab scores appear in tools only when an API key is configured — and never become Index figures.
Limitations
- Duyuz sample, not a census. A percentage here is a share of this sample only.
- Headcount was not verified. Several names are specialty brands or regional chains.
- One homepage per business, no device emulation. Text that exists only after JavaScript runs is invisible here, so a missing H1 can mean the server HTML has no H1.
- Blocked requests, TLS failures, DNS failures, parking pages and responses over 1.2 MB are outside n.
- Core Web Vitals were not measured. This edition has no lab or field CWV scores.
How Website Intelligence feeds the Digital Index
Duyuz Website Intelligence scores individual public sites. Anonymous aggregates from disclosed samples become Digital Index reports and sector benchmarks. We claim the measured n only — currently n=24 on the flagship edition — and grow samples over time. We do not claim “10,000 sites analysed”.
Related
Future editions keep the same honesty pattern: new year, new page, old figures stay as the record of that run.