Age-weighted population demand: counting only the ages that buy what you sell
A headline population treats a toddler and a retiree as the same customer. How to weight each age group by how much it buys from you, and compare areas on the result.
Part of the guide Location intelligence automation
Age-weighted population demand is a count of residents in which each age group is multiplied by how likely it is to buy what you sell, so that the result reflects your customers rather than everyone. A running-shoe shop might count adults in their twenties and thirties in full, older adults at half, and children not at all. Summed over an area, the weighted figure ranks catchments far better than the raw total.
This is for retailers, clinic groups, fitness and education businesses, consumer brands and analysts who compare areas for expansion or sales targets. It shows how to choose the weights, how to compute them from age groups, a worked example where the ranking flips, mistakes to avoid and a flow that applies the weights to every area you care about.
Why the raw count misleads
Two neighbourhoods with 40,000 residents each look equal on a map. One is a student and young-professional district; the other is a settled suburb of retirees. For a nightclub, a protein-shake bar or a hearing clinic, they are completely different markets. The raw count hides that difference, and any decision built on it inherits the error.
Age is not the only thing that drives demand, but it is one of the few you can measure cheaply and consistently across a whole city. A population grid that splits residents into nine ten-year bands, from 0 to 9 up to 80 and over, gives enough detail to build a weighting that fits most consumer businesses.
Choosing the weights
A weight is simply a number between 0 and 1 for each band: 1 for your core buyers, something smaller for occasional ones, 0 for people who never buy. There are three honest ways to set them.
- From your own customers. If your loyalty or booking data records ages, compare the share of each band among customers with its share among residents. Bands that are over-represented get higher weights.
- From industry knowledge. A paediatric dentist knows children are the patients; a retirement-living operator knows the 70s and 80s are the market. Write the weights down and agree them with the team.
- From a test. Start with simple weights, compare the weighted figure with sales at your existing branches, and adjust until the best branches score highest.
A worked example where the ranking flips
Take a made-up fitness studio chain looking at three districts of Melbourne. The weights are its own guess: ages 20 to 29 at 1, 30 to 39 at 1, 40 to 49 at 0.6, 50 to 59 at 0.3, everyone else at 0. The resident figures are invented for illustration.
| District | All residents | Aged 20 to 39 | Aged 40 to 59 | Weighted demand |
|---|---|---|---|---|
| Inner north | 38,000 | 17,500 | 7,200 | about 20,700 |
| Bayside | 46,000 | 9,800 | 13,100 | about 15,700 |
| Outer east | 52,000 | 11,200 | 12,900 | about 17,000 |
Invented figures. Weighted demand assumes half of each 40 to 59 group is 40 to 49, for simplicity.
On raw residents the outer east wins. Weighted, the inner north comes first by a clear margin, because so many of its people fall in the bands the studio sells to. Bayside drops to last. The chain would visit the inner north first and treat the outer east as a second-wave option.
Mistakes with weighting
- Weights that sum to nonsense. Weights are shares of a band, not percentages of the whole. Keep each between 0 and 1.
- Precision you do not have. A weight of 0.37 implies knowledge you rarely possess. Round to tenths and test.
- Forgetting competition. High weighted demand with many rivals may still be a poor bet. Divide by the number of competitors in the same area.
- Applying one set of weights to every product. A pharmacy's prescription business and its cosmetics counter have different buyers. Use separate weights for separate lines.
- Comparing across countries without care. Population figures exist in about 100 countries, such as Australia, Canada, France, India, Thailand and Egypt, but not in the United Kingdom or the United States.
Applying the weights with an AI employee
In xMap Flows you write the weights once and an AI employee applies them to every area on every run. The site-selection side is explained in the guide to AI employees for site selection.
- Area of interest: the districts you compare, drawn or picked by name (Area of interest); or My data with your candidate sites.
- Demographics: totals for each area, or residents within a set distance of each site, with all nine age groups (Demographics).
- Calculated field: one line with your weights, for example demand = round(age_20_29 + age_30_39 + age_40_49 * 0.6 + age_50_59 * 0.3) (Calculated field).
- Nearby places with counts, if you want competitors, and a second line such as demand_per_rival.
- Sort and limit: highest weighted demand first.
- Email or Make a report: the ranked areas with the raw and weighted figures side by side.
The inbox gets a table that shows both numbers, so nobody has to take the weighting on trust. Change a weight and the next run reflects it. The template Score every neighbourhood for a new branch is a good base, Demographic profile of each city gives the age bands for many cities at once, and Where seniors need home care most shows a weighting aimed at the oldest bands. See also the retail and franchise page.
Questions
What age groups are available?
Nine bands of ten years, from 0 to 9 up to 80 and over, plus three summary groups: 0 to 19, 20 to 59 and 60 and over.
Can I weight by income as well as age?
Not from xMap's population grid, which counts residents by age and sex. You can join your own income or customer data to areas with My data and add it in the same Calculated field.
How do I know my weights are right?
Score your existing branches with them. If the best performers come out on top, the weights are doing their job; if not, adjust and test again.
Is weighted demand the same as sales potential?
No. It is a better count of likely buyers. Sales also depend on competition, access, prices and how well a site is run.
Templates that do this
Read next
More on population and demographics · Every article · Templates