How charities find the areas with the fewest services for the people who need them
A charity decides where to send a van, open a hub or fund a partner by counting the people it serves against the services near them. How to do it fairly, district by district.
Part of the guide Workflow automation
A charity finds the areas with the fewest services by splitting its city into equal pieces, counting in each piece the people it exists to help (older residents, young children, everyone) and the services that already help them (food banks, clinics, community centres, pharmacies), and ranking the pieces by people per service. The areas at the top of that list, checked by someone who knows the ground, are where a new hub, a mobile van or a partner grant does the most good.
This is for programme managers, operations leads and grant officers at non-profits, food networks, health charities and community foundations who have to justify where the next pound or dollar goes. It covers which people and which services to count, how to keep the comparison fair, a worked example, the traps that make a well-served area look empty, and an AI employee that keeps the ranking up to date.
Decide who you serve before you count anyone
The commonest mistake is to count everyone. A lunch club for older people cares about residents aged 60 and over; a school uniform bank cares about children; a walk-in clinic cares about everyone but especially the very young and the very old. Demographics gives residents in nine age groups, plus the share of children and of people aged 60 and over, for a circle around a point or for a whole area. Pick the group your service is for and use only that number.
Population figures from xMap's grid exist in some countries only, among them France, Canada, India, Thailand, Egypt and Australia, and not the United Kingdom or the United States. Where there are none, a charity can use its own referral records instead: a table of the postcodes or addresses its clients came from, uploaded to My data, tells you where need already shows up.
Count the services that really do the same job
On the other side of the ratio sit the places that already meet the need. Nearby places searches xMap's places database by type: food banks, soup kitchens, social services organisations, community centres, senior centres, clinics, pharmacies and many more. Two rules keep the count honest.
- Same job, same people. A senior centre and a youth club are both community services, but only one serves your lunch club's members. Keep the types that a person in need could use instead of you.
- Open, not just listed. Ask Nearby places for open places only. A food bank that closed last spring still sits on old directories and makes an area look served.
- Your own sites and partners count too. Add them from My data, so the map does not send you to the street where your own van already stops on Tuesdays.
Keep the comparison fair
People per service is only fair when every piece of the city is the same kind of piece. Districts drawn by a council vary hugely in size; a large district will hold more of everything. The cleaner way is a grid of equal circles or squares, or one circle of the same size around each candidate site. A 1 km circle suits services people walk to; a few kilometres suits services people take a bus to. Whatever you choose, use the same size everywhere.
Then look at more than the ratio. A piece with 12,000 older residents and one community centre scores the same as a piece with 1,200 and none, but they are different problems: the first needs capacity, the second needs presence. Keep both numbers in the table.
A worked example in Lyon
This example is made up. A charity that runs lunch clubs and befriending visits for older people in Lyon wants to place one new weekly club. It lays eight circles of 1 km over the neighbourhoods its volunteers can reach, counts residents aged 60 and over in each, and counts senior centres, community centres and existing lunch clubs inside each circle, including its own two.
| Circle | Residents 60 and over | Services found | People per service |
|---|---|---|---|
| A | 5,400 | 6 | 900 |
| B | 3,900 | 1 | 3,900 |
| C | 7,200 | 3 | 2,400 |
| D | 2,100 | 0 | no service |
| E | 6,600 | 5 | 1,320 |
Made-up figures to show how the comparison reads; three circles with ordinary ratios are left out.
Circle D has no service at all, but only 2,100 older residents; a volunteer who knows the area points out that most of it is a hospital campus and a business park. Circle B has nearly 4,000 older residents and a single community centre, which closes at noon. Circle C has the most people but three services already. The charity opens its club in B, and puts D on a list for a visiting volunteer rather than a weekly club.
Mistakes that hide the real gaps
- Trusting an empty circle. No services can mean no people. Check the population before you celebrate a blank.
- Counting services nobody can use. A private clinic with long waits or a centre for one faith only may not serve the people you help. Read the names of what is inside.
- Forgetting barriers. A river, a motorway or a steep hill can make a service 400 metres away unreachable on foot. Where walking matters, draw a walking area with Travel area instead of a circle.
- One count, never repeated. Services open and close every month. A ranking made once goes stale before the funding round.
Doing it with an AI employee
The count above is a few hours of work by hand and a few minutes for an AI employee, which can then repeat it every month.
- Schedule: the first Monday of each month.
- My data: one row per circle centre or candidate site, with its position.
- Demographics: residents within 1 km of each point, with the age groups.
- Nearby places: set to pass on each incoming place with counts, so every circle gets the number of senior centres, community centres and food banks around it.
- Calculated field: people per service for the age group you serve.
- Sort and limit: highest first, keep the top ten.
- Make a report: a short PDF for the trustees, or an Email to the programme team.
What lands in the inbox each month is the ranked list: every area with its people, its services, the ratio and the change since last month, so a new food bank opening or a centre closing moves an area up or down on its own. Templates to start from: older people far from a community centre, neighbourhoods far from a food bank and needs assessment for a new district. More for the sector on the non-profits page.
Questions
What if my country has no population figures?
Use your own client or referral records instead. Upload where people came from as a table and count clients per service in each area; it shows where need already reaches you.
How big should each area be?
As big as the distance your clients will travel to you. About 1 km for a walk-in service, a few kilometres for one people reach by bus. Keep every area the same size.
Should a charity open where there are no services at all?
Only after checking the people. An area with no services often has few residents, or is a park, campus or industrial zone.
Can the ranking include our partners?
Yes. Add your partners' sites to a table in My data and count them with the other services, so you never plan a hub on top of a partner.