2,375Candidates
2,375Websites
64%URLs responding
441DNS / connection failures
659Redirect elsewhere
April 2025September 2026 — 64% responding
What happened to the sites?
Pick a lane, then drill to the ledger.
One hex, one seat
Colour and dots encode HTTP response rate at the audit, not content survival. Click to peek.
66–100% responding · 40–65% responding ·· 0–39% responding ··· no site (17)
Keyboard: Tab moves to the map, arrow keys move between constituencies, Enter opens one. The full data table below (Browse the ledger) is the accessible alternative.
650 seats · 633 with scraped sites · 17 empty. Methodology ↗ · Automatic Knowledge v5 (OGL) + ONS. Browse ledger ↗
Which parties’ URLs responded?
Click a row to filter the ledger.
- Conservative and Unionist Party65% responding · 569 sites
- Labour Party61% responding · 522 sites
- Green Party80% responding · 301 sites
- Liberal Democrats70% responding · 270 sites
- Independent59% responding · 257 sites
- Reform UK36% responding · 154 sites
- Labour and Co-operative Party58% responding · 108 sites
- Scottish National Party (SNP)33% responding · 39 sites
How the record is made
- Collect. National roster (Democracy Club) + Campaign Lab’s April 2025 crawl of candidate sites.
- Audit. Re-request every URL. “Responding” means a successful HTTP response at the audit. Candidate-name matching is a separate heuristic, not proof of campaign content —Dawn Furness led to a domain-sale page at the audit;Adam Pugh redirected to about.me at the audit.
- Record. Each candidate gets a public name —
p5693.civicord.eth(see record ↗)— website URL, HTTP status and candidate name on Sepolia testnet. Path in Methodology. - Index. Publish a change log so the record can be queried — by researchers, journalists, and agents.
- Verify. No account, no wallet, no permission. Inspect the recorded fields and source evidence separately.