FlowsBook a callSign inCreate free accountSign up
Workflows by industry

How food banks decide where to put a distribution point

A food bank puts pantries where people who need them live and can get to on foot or by bus. How to find those gaps with population, travel time and the help already there.

Part of the guide Workflow automation

By the xMap team · · 5 min read

A food bank should place a distribution point where many households live beyond an easy walk or short bus ride from any existing pantry, in a building a volunteer team can open on time and a van can reach. So the work is three checks in order: where people live, how far they must travel to the help that already exists, and which candidate buildings close that gap. Need indicators from your own referrals sit on top of all three.

This guide is for food banks, community pantries, faith groups running meal programmes and the councils that fund them. It explains how to think about reach rather than distance, a made-up example, the traps that waste a good site, and an AI employee that keeps the picture current every month.

Reach matters more than distance

A pantry two kilometres away on a map can be next door for a family with a car and out of reach for an older person who walks with a trolley. That is why a circle drawn around each pantry flatters the network. The honest measure is travel time on the roads and paths people actually use.

  • Walking reach. Most clients of a neighbourhood pantry arrive on foot. A 15-minute walk covers about a kilometre in a town, less where a motorway or a river splits the streets.
  • Public transport reach. Where people take a bus, the stop matters as much as the building. A site 200 m from a frequent route serves a far wider area than one tucked into a quiet estate.
  • Delivery reach. For the van that brings stock from the warehouse, a 15-minute drive area shows which candidate sites can be restocked within a morning run.

The Travel area block draws exactly this: the area reachable within a travel time or a distance by road, on foot, by bike or by car, following the streets rather than a straight line. Times are for usual traffic, not the moment of the run.

Where the people are

Next, count who lives inside and outside those walking areas. Demographics reads xMap's population grid, small hexagons of about 0.1 km², and returns residents, men and women and nine age groups for any area. Two age groups matter most for food support: children, because families with several children run short first, and people over 60, who travel least.

Residents are only half of need. A dense district of students and a dense district of pensioners need different hours and different food. Your own records of who comes, and from which streets, are the best guide to who is missing.

A made-up example: one city, four pantries

Say a food bank in Montreal runs four pantries and wants to open a fifth. These figures are invented to show the method, not taken from any data.

  1. Draw a 15-minute walking area around each of the four existing pantries.
  2. Draw the city's districts as areas and count residents in each, then count how many live outside every walking area.
  3. The result: one eastern district has about 41,000 residents, of whom some 29,000 live beyond a 15-minute walk of any pantry, and its share of people over 60 is the highest of the city.
  4. List the buildings in that district that could host a weekly session: community centres, places of worship, schools and libraries.
  5. For each, check that it sits inside a 15-minute drive of the warehouse, so the van can restock it, and walk the nearest bus stop yourself.

In this example two halls pass every check. The food bank pilots one on Saturday mornings, counts who comes and from where, and keeps the second as a back-up if the first fills within a month.

Mistakes to avoid

  • Counting buildings, not reach. A district can have three pantries all on one side of a rail line. Count people outside walking areas, not pantries per district.
  • Ignoring other providers. Soup kitchens, school breakfast clubs and church larders also feed people. Map them too, or you will open next to help that already exists.
  • Choosing a hall nobody can open. A perfect location fails if no volunteer lives near enough to unlock it at eight on a Saturday. Check your volunteer list against the site.
  • Planning once. New housing, a closed community centre or a changed bus route moves the gaps. Look again every few months.

Doing it with an AI employee

An AI employee in xMap Flows can redo the analysis every month and tell the team what moved. Built from blocks, in this order:

  1. Schedule: the first Monday of every month.
  2. My data: your pantries, uploaded once as a table with their addresses.
  3. Travel area: a 15-minute walk around each pantry.
  4. Nearby places: community centres, places of worship and libraries in the area you serve, the candidate hosts.
  5. Demographics: the people within 1 km of each candidate, with children and seniors.
  6. xMap LLM: rank the candidates by people within reach who live outside every walking area, and write one line of reasoning each.
  7. Email and Show on a map: the ranked list and the map to the operations lead.

What lands in the inbox on the first Monday of the month is a short ranked list of buildings with the people around each and why it was ranked there, plus a map showing the walking areas and the gaps between them. The template Neighbourhoods far from a food bank starts from much of this, and Routes for the weekly surplus food pickup plans the collection side once the sites are chosen.

Questions

How far should people have to walk to a food bank?

There is no single rule, but a 15-minute walk, about a kilometre in a town, is a common planning line. For older clients and families with small children, aim shorter.

Can I do this without population data for my country?

Yes. Use your own referral addresses as the map of need, and draw walking areas around your pantries to see which referred households live outside them.

Does the travel time include waiting for a bus?

No. Travel area works on foot, by bike or by car along the roads. For bus riders, check the distance from each candidate site to the nearest frequent stop yourself.

How often should a food bank redo this?

Monthly is enough for most. A scheduled AI employee repeats it and only the changes need a person's attention.

Templates that do this

Read next

More on workflows by industry · Every article · Templates

Put it to work

An AI employee in xMap Flows does this on live data, on its own timetable. Start from a template or describe it in plain words.