Charmgen field guide
How to compare country data without fooling yourself
Country tables look objective, but a clean row of numbers can hide incompatible years, units and concepts. A useful comparison begins by deciding what question the data should answer, then checking whether each indicator is capable of answering it.
01
Start with the decision, not the available column
“Which market is bigger?” can mean total output, number of residents, consumer spending or addressable online users. GDP answers the production question; population answers scale; internet use only approximates digital reach. Choose the concept before looking at which country ranks first.
Use several indicators when the decision has several dimensions. A large population with low income per person can support a different product strategy from a small, wealthy and highly connected market. Combining the measures is analysis; collapsing them into one unexplained score is usually not.
02
Align years and preserve missing values
International datasets update on different schedules. One country may have a 2025 GDP observation and a 2023 education observation. Comparing both as if they describe the same moment creates false precision, particularly around recessions, wars or policy shocks.
A blank observation is information about the dataset, not evidence that the value was zero. Keep it blank, show the latest available year and decide whether the gap makes the comparison unusable. Interpolation can be appropriate in formal analysis, but it must be explicit.
03
Separate scale, rate and level
Total GDP measures scale, GDP growth measures year-to-year movement, and GDP per capita divides output by population. A large economy can grow slowly; a smaller economy can grow quickly from a lower base. None of those facts contradicts the others.
Percentages also need denominators. Trade at 100% of GDP does not mean every unit of domestic output is exported: trade is a gross flow while GDP is value added. Internet subscriptions per 100 people can exceed 100 because people and businesses can hold multiple subscriptions.
04
Finish with a source check
Read the official definition and original data organisation. Modeled ILO unemployment is designed for cross-country comparison but may differ from a national headline release. Urban population depends on national definitions of “urban.” These differences are not footnotes; they shape what the number means.
Treat the first comparison as a map of questions to investigate. For a consequential decision, continue to national statistical agencies, sector data, regulation and city-level conditions. A country average is a starting point, never a complete market.