How We Verify Every Data Point

We wrote recently about why sourced, verified data matters when choosing where to live and work abroad. The argument is straightforward: a number without provenance is just a guess in disguise. But principles are easy to state and harder to practice. So here's how we actually do it -- the specific sources, standards, and processes behind every metric on Harborage.

The verification standard

Every data point on Harborage must satisfy three requirements before it reaches you:

1. Traceable to an authoritative source. Not "user-submitted" or "estimated." A specific organization, dataset, or measurement system with a known methodology. When you see a number on Harborage, you can follow it back to where it originated.

2. Dated and scoped. Every metric carries a timestamp showing when it was last verified and a geographic scope showing exactly what it measures. "Average rent in Lisbon" is not the same as "median rent for a furnished one-bedroom in Lisbon's Santos neighborhood." We show you which one you're looking at.

3. Transparent about limitations. No dataset is perfect. Some metrics are country-level when we'd prefer city-level. Some are updated annually when we'd prefer monthly. We tell you these limitations rather than hiding them behind a polished interface.

If a metric can't meet all three requirements, it doesn't ship. We'd rather show you fewer numbers you can trust than more numbers you can't.

Our data sources

Here's what powers Harborage today, and what we're integrating next.

Internet speed: Ookla Speedtest Intelligence

When a nomad platform tells you a city has "fast internet," what does that actually mean? Usually, it means some users ran speed tests and the platform averaged the results. That approach has problems: the sample is self-selected (people test their speed when it feels slow or when it feels fast, not randomly), and it tells you nothing about variation across neighborhoods or times of day.

We use Ookla's Speedtest Intelligence dataset, which aggregates billions of tests across 190+ countries. Ookla's data is the industry standard for internet performance measurement -- used by telecommunications regulators, the FCC, and the European Commission. More importantly for our purposes, Ookla provides data at the tile level: roughly 610-meter squares that let us measure internet speed not just at the city level but at the neighborhood level.

For our current 10 cities, we show verified download and upload speeds with the measurement methodology clearly attributed. As we expand, Ookla's tile-level granularity will power our neighborhood internet maps -- so you can see that Lisbon's Príncipe Real averages 120 Mbps while Alfama averages 45 Mbps, rather than a single "Lisbon: 85 Mbps" that helps no one choose where to rent.

Safety: Global Peace Index and OSAC

Safety is arguably the highest-stakes metric we publish. Get it wrong and the consequences aren't financial -- they're personal.

We currently draw safety scores from the Global Peace Index, which evaluates 163 countries across 23 quantitative and qualitative indicators compiled by the Institute for Economics and Peace. It's an imperfect tool -- country-level data obscures city-level variation, and a composite index necessarily makes judgment calls about how to weight different risk factors.

That's why we're building a second safety layer using OSAC (Overseas Security Advisory Council) reports. OSAC is run by the U.S. State Department in partnership with private sector security professionals. Their country and city assessments cover crime, terrorism, political violence, civil unrest, transportation safety, and health risks -- with narrative detail that we extract into structured ratings.

The combination matters. The Global Peace Index gives us a quantitative baseline. OSAC gives us qualitative depth and city-specific context. Together, they produce safety assessments that are more nuanced than either source alone.

Our long-term goal is neighborhood-level safety data. We're exploring partnerships with providers like GeoSure and investigating local crime statistics for our top cities. Until we can verify safety at the neighborhood level, we'll be transparent about the limitations of country and city-level data.

Cost of living: World Bank and verified local data

Cost-of-living data is where the gap between perception and reality is widest. Most nomad platforms rely heavily on crowdsourced cost submissions, which skew toward certain neighborhoods, demographics, and spending patterns. The person submitting costs from a serviced apartment in Barcelona's Eixample lives a different financial reality than someone in a local flat in Gràcia.

We use World Bank Open Data for our macroeconomic baseline: CPI, inflation rates, GDP per capita (PPP-adjusted), and the Gini coefficient for income inequality. These numbers come from national statistics offices and international surveys with documented methodologies -- and they're freely available under CC BY 4.0.

For city-level cost breakdowns (rent, food, transport, utilities, entertainment), we currently use aggregated data from established indices, clearly attributed. We're not satisfied with this yet. Our roadmap includes integrating more granular, verified cost data as we expand to 50 cities. Every cost figure will carry its source, date, and geographic scope so you can evaluate whether it reflects your situation.

Climate: NOAA

When you're planning a move, "nice weather" isn't useful. You need to know: What's the temperature range in September? How much does it rain in October? Will humidity make the heat unbearable in July?

We're integrating NOAA (National Oceanic and Atmospheric Administration) data for climate metrics. NOAA operates over 100,000 weather stations across 180 countries, and their 30-year Climate Normals provide statistically robust baselines for temperature, precipitation, humidity, and wind.

Station-level data is particularly valuable because it lets us show climate variation within a city. Bangkok's riverside neighborhoods and its elevated areas experience meaningfully different microclimates. NOAA's station density in urban areas often reaches neighborhood-level resolution.

Our current temperature data provides a starting point. The full NOAA integration will deliver monthly climate profiles for each city -- so you can plan your move around actual weather patterns, not vacation-brochure generalizations.

Air quality: IQAir

Air quality is a metric that most nomad platforms either ignore entirely or show as a current snapshot. Neither approach serves someone making a multi-month relocation decision. You need seasonal patterns: Bangkok's air quality in March is radically different from Bangkok's air quality in August. If you have children or respiratory conditions, this isn't a nice-to-know -- it's a dealbreaker.

We're integrating IQAir data for real-time and historical air quality metrics. IQAir monitors AQI, PM2.5, and other pollutants across 5,000+ cities using a network of government monitors and validated sensors. Their data is used by the UN, WHO, and Greenpeace for global air quality assessment.

Our integration will show current AQI, seasonal trends, and health-relevant breakdowns -- with special attention to metrics that matter for families with young children. Because a city that looks great on paper in December might be genuinely hazardous in burning season.

Healthcare: WHO Healthcare Access and Quality Index

For families especially, healthcare access is non-negotiable. But "good healthcare" means different things to different people. A solo nomad with no chronic conditions needs emergency care access. A family with young children needs pediatric specialists, vaccination availability, and English-speaking doctors.

We're integrating the WHO Healthcare Access and Quality Index (HAQ), which scores 204 countries on a 0-100 scale based on mortality from 32 causes that should be preventable with effective healthcare. The index comes from the Global Burden of Disease study at the Institute for Health Metrics and Evaluation -- the most comprehensive epidemiological dataset in existence.

The HAQ score alone doesn't tell the whole story. That's why we're supplementing it with JCI (Joint Commission International) accreditation data, which maps internationally accredited hospitals to specific locations. Combined with HAQ, this lets us show not just "how good is healthcare in this country" but "how far are you from a hospital that meets international standards."

What we don't do

Transparency about methodology means being honest about what we don't do, not just what we do.

We don't use unverified self-reported data as a primary source. Community submissions can be valuable for surfacing patterns and validating other sources, but they don't meet our standard for primary metrics. If the only source for a data point is "some users said so," that data point doesn't appear on Harborage.

We don't present country-level data as city-level data. Many metrics are only available at the country level -- healthcare quality, economic indicators, some safety data. When we use country-level data, we label it as such. We'd rather show you an honest "country-level" tag than quietly pretend the data is more specific than it is.

We don't hide stale data. When a metric hasn't been verified in over twelve months, we flag it. Stale data can still be useful context, but only if you know it's stale.

Where we're headed

We currently cover 10 cities with verified data across five categories: internet speed, safety, cost of living, climate, and walkability. That's a starting point, not a destination.

Over the coming months, we're expanding to 50 cities with deeper data coverage:

• Automated data pipelines pulling from World Bank, NOAA, IQAir, and M-Lab APIs to keep metrics fresh without manual intervention
• Neighborhood-level internet data using Ookla's 610-meter tile resolution for our top 20 cities
• OSAC safety assessments structured from narrative reports for all 50 cities
• Family-specific metrics including school access, pediatric healthcare, and child safety indices
• Seasonal climate and air quality profiles so you can plan around actual conditions, not annual averages

Every expansion follows the same standard: traceable, dated, scoped, and transparent about limitations.

Verify us

We built Harborage on the principle that data should be checkable. That applies to our own data too. Every metric on our platform links to its source. If you find a discrepancy between our published figure and the underlying source, we want to hear about it.

Check our full methodology for detailed documentation of each data source, its update frequency, and its limitations. Explore our cities to see source-linked data in practice.

The world is full of platforms that ask you to trust their numbers. We'd rather show you ours and let you decide.