Resources
Data Sources
Where the figures come from, how they are produced, and how much weight to put on them.
Read this first
Urblytica's city figures are researched estimates. For each city, an AI research pipeline searches the live web, reads what it finds, and extracts the numbers into our database. The public sources listed below are what that search returns and what the figures are drawn from — they are references, not feeds we ingest.
So the number stored against a city is our pipeline's reading of a public source, not a row copied from that source. We do not hold licensed data subscriptions, and we are not going to imply otherwise. Where a reading comes back implausible it is discarded rather than scored, and where a city is missing a component the score says so.
Every city page records which research pass produced its figures, the internal quality score of that pass, and the date it was last refreshed — and shows the component-by-component arithmetic behind its score. See the methodology for the weights and the accepted ranges.
For a decision that turns on an exact figure — a mortgage, a relocation package, a lease — check that figure against the primary source. Urblytica is built to tell you which cities are worth that effort.
The sources our research draws on
National census & statistical offices
Population, demographics, median household income, education and housing characteristics. In the United States this is the Census Bureau's American Community Survey; elsewhere it is the equivalent national statistical office.
Labour-market statistics
Wages, unemployment rates and industry employment. In the United States, Bureau of Labor Statistics releases; elsewhere, the national labour or employment ministry.
Economic output reporting
Metropolitan and regional GDP, personal income and growth figures, as published by national economic analysis bodies.
Published crime statistics
City-level crime reporting, which feeds the safety component. Coverage and comparability vary a great deal by country, and in places where a city publishes nothing usable the safety reading is left missing rather than guessed.
Climate records
Long-run temperature, precipitation, snowfall and sunshine averages from national meteorological services.
Rental & home-price reporting
Published rental and home-price levels from property-market sources, used for the housing part of the cost of living.
Transit & walkability reporting
Public-transport coverage and walkability, as described by local transit authorities and published assessments. We do not license a proprietary walk score.
Company registries & filing fees
Entity filing fees, processing times and requirements, taken from the registry of each state or country for the Business Formation module.
Care-facility inspection records
Nursing-home inspection results, staffing and quality ratings behind Senior Care Intelligence — in the United States, the CMS Care Compare dataset.
Business directory listings
Business profiles, reviews and claims are submitted by owners and visitors, then moderated. They are first-party content, not researched figures, and are not part of any score.
How often this is refreshed
A city's figures are researched once and then re-researched when they go stale; the date of the last pass is stored per city and printed on the city page, so you never have to guess how old a number is. Scores are recomputed nightly from whatever figures are current, and in full whenever a weight on the methodology page changes.
Refresh rates are not uniform. Cities with visitors get refreshed sooner, and a city nobody has opened may be carrying its original figures. The date on the page is the honest answer for that city.
Found a figure that looks wrong?
Tell us which city and which figure. The breakdown table on each city page shows exactly which reading drives its score, which makes a report easy to act on — and a bad reading corrected once is corrected for everyone.
