Guide

How to read a demographic report

Most demographic reports are read badly: people scan for the population number, quote it, and ignore the twenty numbers that actually decide whether a site works. This walkthrough goes through a real report — our public sample — section by section, and shows what each part is for.

The short answer

Read a demographic report in four passes: check the catchment definition first, then the population base, then the composition (age, income, households), and only then the index against the regional and national benchmark. The index is what tells you whether a location is unusual; the raw counts only tell you whether it is big.

1. Start with the catchment definition, not the numbers

Every figure in the report is conditional on the catchment. A 3 km radius around a CBD address and a 3 km radius around a highway interchange are not comparable populations, and neither is a 3 km radius compared against someone else's 10-minute drive time.

Before reading a single metric, confirm three things: the centre point, the catchment method and size, and the census vintage. If a report does not state all three on the front page, you cannot audit any number in it.

  • Centre point — the exact address or coordinate the catchment is drawn from.
  • Method and size — radius in kilometres, or travel time and mode.
  • Census vintage — which census year the underlying counts come from.

2. Read the population base before the composition

Total resident population sets the ceiling on everything downstream. A catchment with an exceptional income profile and 4,000 residents is usually a worse retail prospect than an average catchment with 60,000.

Population also tells you whether the rest of the report is statistically meaningful. Small catchments produce volatile percentages: one apartment block can move a household-composition share by several points. Treat composition percentages in very small catchments as directional, not precise.

3. Read composition as a shape, not a list

Age, income, household composition and dwelling type describe the same population from four angles, and they should tell a consistent story. Young age profile plus small household size plus high rental share plus high apartment share is a coherent inner-urban picture. If those four disagree, something in the catchment is mixed — often a boundary running through two very different neighbourhoods — and you should look at a smaller radius.

Read income as a distribution, not just a median. Two catchments with the same median household income behave completely differently if one is tightly clustered and the other is barbelled between low and high income bands.

4. The index against the benchmark is the actual insight

Every metric in the report is shown alongside the regional and national figure. That comparison is the part a scraper site cannot give you, and it is where site-selection decisions are made.

A catchment with 22% of households in the top income band sounds strong. If the national figure is 24%, it is slightly below average. If the national figure is 11%, it is exceptional. The absolute number means nothing until you index it.

Practical rule: anything within a few points of the benchmark is 'normal for this market' and should not drive a decision. Look for the two or three metrics where the catchment is dramatically off-benchmark — those are the reasons this location is different.

5. Check what is missing before you quote it

Reports flag any section the census data provider could not return for that catchment, and partial reports carry a banner and a pro-rated charge. Missing does not mean zero — a suppressed or unavailable metric is left blank rather than being rendered as 0, because a false zero is far more damaging in a lease paper than a gap.

Similarly, monetary values are shown in the local currency of the data (AUD for Australian data, USD for US data, and so on), which is not necessarily the currency you were billed in.

6. What to actually put in the paper

For a lease, credit or investment paper, the defensible extract is: catchment definition, total population, median household income, the two or three most off-benchmark metrics with their benchmark values, and the census vintage. Everything else is supporting evidence.

Export the full report as PDF or CSV so the recipient can restate any figure. A number nobody can reproduce is a number that gets argued with.

Frequently asked

What is the most important number in a demographic report?

There isn't one. Total population sets the scale, but the decision-relevant information is the index — how far each metric sits from the regional and national benchmark. A metric that matches the benchmark tells you nothing about why this site is different.

Why are some percentages missing rather than zero?

Because a genuine zero and 'no data returned for this catchment' mean completely different things. Blank cells indicate the metric was unavailable or suppressed for that catchment; a zero means the census actually counted none.

What currency are the income figures in?

The local currency of the census data — Australian data is in AUD, US data in USD, Mexican data in MXN, and so on. That is independent of the currency you were charged in for the report.

How small can a catchment be before the percentages stop being reliable?

There is no hard cut-off, but below a few thousand residents individual buildings start to move composition shares by several percentage points. Treat small-catchment percentages as directional and cross-check them against a slightly larger radius.

Run the numbers on your own address

Drop a pin anywhere in six countries and get a census-grade catchment report, benchmarked against the region and the country.