Safe Trip Index

Travellers making real decisions — 'is it safe to go to Thailand right now?' — have to cross-reference half a dozen government advisories (FCDO, State Dept, Smartraveller), disease trackers (WHO, ECDC), conflict monitors (ACLED), and disaster feeds (USGS, NOAA, GDACS). Each source uses a different scale, different update cadence, different definition of 'risk'. The picture you need takes 45 minutes to assemble.

Existing travel-safety sites either repackage a single government source (inheriting its biases), sell fear with red maps and alarm language, or bury the methodology so deep you can't audit the score. None of them update daily from primary sources across all six dimensions.

The brief: build a calm, auditable, daily-updated safety index that synthesises all the primary sources into a single 0-100 score per country, broken into six readable dimensions, with full source transparency and an editorial firewall between scoring and revenue.

The scoring algorithm (v2.3) ingests data from WHO, ECDC, ACLED, USGS, NOAA, GDACS, FCDO, US State Department, Smartraveller, and others — 663 active drivers across every scored country. Each driver maps to one of six dimensions: Disease, Conflict, Disaster, Crime, Unrest, Infrastructure. The composite score is a weighted blend that changes only when the underlying signals change, not on editorial whim.

Every score component cites its underlying signals. Algorithm changes are versioned and announced 14 days ahead for material moves. No state actor can lower a country's score by calling us — their objection is published alongside. Corrections are public within 2 hours of identification. This isn't just editorial policy; it's the product.

The design language is deliberately calm. No fire-alarm red, even at high risk. Colours are warm cream, sage, and muted amber. Headlines are matter-of-fact. The audience pays for calm, not panic — a design principle that extends from the palette to the copy to the data presentation.

The data pipeline runs daily: ingest from each primary source, normalise to the internal schema, compute per-dimension scores, blend into the composite, diff against yesterday, and publish. The diff surfaces as the 7-day movement indicator on every country card — 'Portugal 86, up 0 in 7 days' — so users see momentum, not just a snapshot.

The front-end is Next.js with SSG for every country and region page, revalidated daily after the pipeline completes. The globe visualisation uses WebGL — each dot is a scored country, the halo brightness reflects data freshness. Country cards show the composite score plus a six-cell dimension breakdown with individual scores.

Revenue streams are cleanly separated from scoring: Pro consumer subscriptions (€4-9/mo), Safe Trip for Teams (€99-2,000/mo), public API tier, and clearly-labelled travel-insurance and eSIM affiliate links. None of these streams influence the underlying score. A wall between commercial and editorial is published on the About page.