Organizing neighborhood mutual aid that is still running next year

Mutual aid covers several different things, and groups that do not name which one they are doing end up doing none of them well.

Some are preparedness groups: the point is to know, before a storm or a fire, who has equipment and skills and who will need help evacuating. Some are acute-need groups: a neighbor has a new baby, a diagnosis, a death in the family, and the group organizes around that specific situation for weeks. Some are ongoing-exchange groups: rides, childcare, tools, the things people need every week and mostly do not ask for because asking feels like an imposition.

Most successful groups do all three eventually, but they start with one and add the others once the habit exists. Trying to launch all three at once produces a structure nobody uses because no single part of it ever got busy enough to become normal.

Be equally clear about what the group is not. Mutual aid is not a charity, it is not a substitute for emergency services, and it is not a referral service for professional care. Groups that stay inside their scope last. Groups that drift into being everyone's backstop exhaust their volunteers.

A group chat's membership is whoever has not muted it. That is not a roster, and every piece of coordination downstream inherits the ambiguity.

Member Directory builds a roster with contact details, role, and active status, including entries for people who belong in the directory without being app members, which matters here more than in most contexts. The neighbor who is eighty-two and the reason the group exists will not install anything, and they still need to be in the list.

Two habits make the roster useful rather than decorative. Record who is a point of contact for whom, so the group knows which people have someone checking on them and which do not. And review it after any turnover on the street, because an outdated roster produces the specific failure mutual aid exists to prevent, which is a person nobody called.

The most valuable thing a preparedness group holds is boring on a normal day and decisive on a bad one: an inventory of what people can offer.

Neighborhood Emergency keeps a skills and equipment registry that each neighbor maintains themselves, a phone tree of block captains with the area each covers, and a storm-mode check-in broadcast. In storm mode a coordinator sends an "are you OK?" push, neighbors answer with one tap, and coordinators see the board of who is accounted for. Broadcasts are restricted to a coordinator group, enforced on the server, which is the detail that keeps the tool usable: if anyone could send an all-neighbors alert, people would mute it within a month and it would be muted on the day it mattered.

1. Ask everyone to list what they have and what they can do, and say plainly that a chainsaw, a truck, a spare room, and nursing training are all on the same list. People under-report because they assume only dramatic capabilities count.
2. Record who needs help rather than only who can give it. This is the half groups omit, and it is the half that determines whether the check-in produces a rescue or a headcount.
3. Divide the area between block captains so every household is somebody's responsibility, with no gaps and a deliberate overlap at the edges.
4. Run a check-in once when nothing is happening. The first real use of an untested broadcast is not the time to discover a third of the group never enabled notifications.
5. Review the registry annually, ideally in the month before whatever season you are preparing for.

The hardest problem in mutual aid is not recruiting helpers. It is that the people who need help most ask least, because a request in a group chat feels like standing up in a room and saying you are struggling.

The fix is structural rather than emotional. When there is an existing sheet with empty slots on it, taking a slot is participation rather than charity, and adding a need to a list is administration rather than confession.

Volunteer runs sign-up sheets with capacity-limited slots that members claim, swap, and release, with capacity enforced by the server so a slot is never double-booked even when two people tap at once. Use it for anything recurring: the ride rota to dialysis, the school-run backup, snow clearing for the three houses that cannot do their own, the two people who check the storm drains before heavy rain.

The swap-and-release behaviour matters more here than the sign-up. Volunteers do not usually vanish because they stopped caring. They vanish because they could not make one week, had no way to hand it off, felt bad, and quietly stopped answering. A release button is a retention feature.

When a specific household needs concentrated help for a few weeks, the people who show up are mostly not in the group. They are coworkers, church, the school-gate crowd, family two towns over. None of them will install an app to bring a lasagne.

Meal Train is built around exactly that gap: you set up the care period, lay out the nights, note dietary needs and how to deliver, and share one link. People claim a night and write in what they are bringing, with no account and no sign-up. Claims are handled by the server, so two people cannot take the same night, and sign-ups collected through the link cannot be edited or deleted from inside the app, only viewed.

Some practical notes that decide whether one of these actually helps the family.

1. Appoint one coordinator who is not the affected household. The single most common mistake is leaving the person in crisis to run their own support rota.
2. Put the delivery instruction on the train itself, in plain words. "Cooler on the porch, please do not ring the bell, the baby sleeps at seven" prevents more distress than any amount of good intent.
3. Schedule fewer nights than people offer. Four meals a week for six weeks helps more than fourteen meals in nine days followed by silence.
4. Extend it rather than over-planning it. The hardest stretch is usually week three onward, after the initial rush, and a train that can be extended reaches that far.

Preparedness and crisis response are the visible parts. The exchange that keeps a group alive between them is small, constant, and should not route through a person.

Marketplace is a classifieds board scoped to the community, with listings, photos, categories, questions to the seller, and moderation of flagged items by leadership. The give-away half is the mutual aid half. A cot, a wheelchair ramp, a bag of school uniform in the next size up.

Lost & Found takes posts for lost or found items with a photo and last-known location, suggests likely matches by category and keyword, and closes both posts when an item is claimed. It is a modest tool that resolves one of the highest-frequency neighborhood interactions there is, and it does it without anyone having to relay messages.

Babysitting Co-op handles the exchange that otherwise dies of accounting. Members log the hours they sat, the other parent confirms, and only confirmed entries count toward a balance, so nobody can mint credit for themselves. There is a coverage board where a member posts a request and one other member claims it. The confirmation step is what makes it durable. Informal childcare swaps collapse when one household starts to feel it is owed more than it is getting, and a ledger both parties agreed to prevents that feeling from being a private grievance.

A group whose only messages are alerts becomes a group people mute.

Announcements posts news with admin approval before publishing, an expiry date, notifications to the audience you choose, and group targeting so something can go to the block captains rather than to everyone. Expiry is the feature to lean on here, because a board of stale notices is how a neighborhood group stops being read.

A rhythm that works for most groups: something monthly whether or not anything happened, a seasonal note before the weather that concerns your area, and a short summary after any real event saying what was covered and by whom. That last one is the one groups skip, and it is the one that produces next year's volunteers. People sign up again for something they saw work.

Every long-running mutual aid group has two or three people doing most of the coordinating, and the group's lifespan is usually the same as theirs.

Three things extend it. Write down what the coordinator role actually involves, in a paragraph, so it can be handed over rather than absorbed. Have at least two people in every coordinating position, even where one would do, so nobody is the single point of failure for the street. And make the shared lists genuinely shared, so the registry, the roster, the rota, and the ledger are not reconstructable only from one person's phone.

The test is simple and worth applying honestly. If the person who started the group moved away next month, would it still be running in a year? If the answer is no, the work to do is not recruitment. It is writing down what only they currently know.

A group like this asks less of everyone than a group chat does. Neighbors are not scrolling to find out whether something still needs doing. The person who needs a ride is picking an open slot instead of composing a request. The coordinator is not the database.

And the thing it actually protects against: the gap between an emergency happening and anyone knowing who is affected. A street with a current roster, a registry of what people have, and a check-in that has been tested once closes that gap in an hour. A street with a group chat closes it in a day and a half, and finds out afterwards who it missed.

• Neighborhood Ready
• Meal Train
• Volunteer Sign-Up
• Lost & Found
• Marketplace
• Babysitting Co-op Ledger
• Member Directory
• Announcements

Every app in the library is included with the household plan. Install the ones this guide uses and skip the rest.

• See the whole organizer toolkit
• How siblings can coordinate care for an aging parent without a group text
• Collecting HOA dues without chasing people every quarter