What this is, and who it is for.
Purpose Crew is a stewarded path back into community for people who have been homeless, out of work for years, or a long time away. This page walks you through the whole of it, at your pace. It is written for the people who could bring it to life: a trusted local organisation willing to be its home, a council or a Crown agency, a business that opens a door, a funder who backs the first one.
It is not written for the people a crew is for. Places on a crew are never advertised or offered here. They are arranged quietly and locally, by a host organisation, once a chapter has formed. If you are here because you need support today, please reach a live service such as your local City Mission, or call or text 1737 any time to talk with a trained counsellor.
We are forming our first crews now, in Auckland to start, and we talk in the order people put their hand up. Step forward early and you are not a name on a waiting list, you are one of the founding few who shape how this runs. The only thing in short supply is founding host and partner slots, never a place for a person who needs one.
And if you are a marae, an iwi organisation or an urban Māori service, you are not a venue we slot into a template. You would be a host who runs this your own way, governs it your own way, and gifts it its name. Our handbook is a starting point, not a rulebook.
One last word, on how we tell this story. Throughout, you will see the work, the places, the tools and the shared table, and you will never see a person at their lowest. Nobody is named or photographed without their say-so. That is not a presentation choice, it is the first thing the model protects.
A few hours of useful work, a shared experience worth showing up for.
The hardest thing to rebuild, after someone has been homeless or a long time out of work, is not income. It is purpose, routine, and the feeling of belonging somewhere. The path back is rarely missing because a person stopped trying. It is missing because no one was holding it open. Purpose Crew exists to hold it open.
The loop is simple, and it is the whole engine.
A few hours of useful work
A small crew, led by a professional, gives about three hours to something genuinely useful nearby: a community garden, a beach clean, an op shop sorted.
A token for showing up
Everyone who turns up earns the same single token. No ranking, no scoring, output is never measured.
A shared experience
The crew spends its tokens together on something worth showing up for, chosen as a group: a night at the pictures, a market, ten-pin bowling, a climb.
The work is real but modest. The crews are small. The pace is gentle. Over weeks, people build routine, trust, and a circle who expect to see them. That is the product: a way back into the rhythm of community, stewarded rather than left to chance.
What this is not.
- Not work-for-the-dole. Taking part is voluntary, never tied to a benefit obligation, never enforced by a sanction.
- Not cheap labour. Tasks are chosen for the crew's benefit, and a written rule means no crew ever does work that someone is, or recently was, paid to do.
- Not a job agency. Work may follow for some, but routine, belonging and readiness are the measures, and staying for years as a steady crew member is success too.
- Not a substitute for housing or health. Those come first and are never judged here. Purpose Crew adds purpose and company on top of them.
What it looks like for a person.
Tama, 43, has slept in his van in Henderson for eight months. He says no twice, comes "just to look", stacks timber for an hour in his first week and leaves without speaking. In week three he misses a session, and nobody chases him. By week six, eight of them see a film together. He laughs twice, and takes the tea run the following week.
Lorraine, 58, has not worked since a redundancy nine years ago and tells the facilitator she is "past being useful". By the end of the term she is the one showing new people where the gloves are kept, and reminding the crew whose turn it is to choose the next outing.
Every crew is led by a professional whose job is the people, not the productivity. That one difference, someone whose work is to hold the day steady and notice who is there, is what separates a crew that lasts from a roster that burns people out.
One line sits under all of it: taking part can never quietly cost a participant money. The tokens never turn into cash, and the whole model is built, and legally checked, so that showing up never puts a person's benefit or their standing at risk. A missed session is never punished either. No token is taken away, no explanation is asked for, just "good to see you" the next time. The crew being there every week, no matter what, is how the trust is built.
Three things people ask straight away, answered plainly.
"Isn't this making homeless people work for movie tickets?" No. It is voluntary, no task displaces a paid job, and the reward is a shared experience the crew chooses, not a wage dressed up as one.
"Why not just give people cash?" For income, cash is better, and we say so. The token is not about money. It is about doing something together that money alone does not buy: belonging.
"Isn't this just a job scheme?" No. Nobody is placed, pushed, or measured on output. Some people move toward work, many simply find a week worth showing up for, and both are wins.
Why it works, and where it fits.
None of this is invented from scratch. Each part stands on tested ground, most of it already running in New Zealand. Every figure here is a planning estimate to re-check before it is quoted to anyone.
- The Clubhouse modelRunning since 1948, and live in Auckland today as Crossroads Clubhouse, run by the Arahura Charitable Trust. Strong, long proof that voluntary, meaningful, unpaid contribution restores purpose and belonging, and that the shape runs legally here. Purpose Crew adds the shared-experience token.
- Housing FirstIn New Zealand, The People's Project in Hamilton. The settled foundation that stabilising people first, unconditionally, is what works. Purpose Crew sits on top of that, never in place of it, and taking part is never conditional on housing status.
- Contingency managementA 2021 JAMA Psychiatry review (Bolívar and colleagues) found that small, immediate, non-cash rewards reliably lift attendance and engagement in hard-to-reach groups. One honest caveat: that research is in clinical addiction care, not homelessness, so we cite it for the transferable mechanism, not as a like-for-like outcome.
- A supervised-crew model already runs hereCultivate Christchurch has run a supervised-crew model since 2014, with young people not in education, employment or training. An adjacent precedent rather than a direct match, and proof the shape works in New Zealand. Confirm its figures with the organisation before quoting.
The need, and the gap.
Homelessness-support funding fell by an estimated $79 million year on year, the year to June 2026 against the year to June 2025, while Auckland rough sleepers were reported to have more than doubled, to around 940 in the year to September 2025. Both figures are attributed to their sources and should be re-checked before quoting. Central funding also arrives in volatile yearly tranches, so a stretched provider cannot plan far ahead. Purpose Crew is low-cost capacity that is held locally and does not depend on those tranches.
The gap it fills is specific: nothing widely deployed combines professional-led, small-group, supervised contribution with a non-cash token spent mostly on shared experiences, aimed squarely at re-initiation with society. And it sits deliberately upstream. It is the lowest-threshold rung that feeds into paid pathways such as Lifewise Works, never a competitor to them.
On the obvious question, why a token and not cash: for income, cash is better, and we say so. The token is a different tool doing a job cash cannot, lifting the simple act of showing up, and keeping the whole thing outside the wage and benefit traps. That is not a hunch, it is the contingency-management finding above.
What "working" means here, set before any money is discussed. Success is measured as routine first, then belonging, then readiness, never as a count of job placements. A person who reaches a comfortable, steady plateau and stays a valued part of the crew is a success, not a failure to launch. Naming this now is what stops the model drifting toward the discredited work-first shape later.
Council or Crown agency? Here is what changes
Seen from inside government
The same model, from where you sit. This is an open door, not a pitch. If your agency's people need a way back to routine and belonging, here is where it fits, and you are the one who knows which of these doors is yours, and what could sit behind it.
Corrections
The Six Pillars of Reintegration are your own framework, and Purpose Crew delivers three of the six directly: meaningful activity, community connection, and wellbeing. It sits naturally alongside Out-of-Gate support in the weeks after release, when routine and belonging often decide whether someone slips back. A host you already contract carries vetting and insurance, so the path is shorter.
MSD and Work and Income
This is upstream work-readiness, the rung below a work programme, for people not yet ready for one. There is a live precedent for funding community activity of this kind in the Activity in the Community participation scheme. A case manager gains a low-threshold place to refer someone who needs routine before a job search means anything.
Kāinga Ora
A way to reach tenants who have gone quiet, the ones a tenancy team worries about but cannot easily help. A crew gives an isolated tenant a reason to leave the flat, a piece of useful work, and a circle who expect to see them next week.
Health New Zealand
Purpose Crew stays firmly in the non-clinical lane. It never assesses, treats or counsels. It complements community mental health and addiction services as a place for people stable enough for group activity, with any clinical concern going straight back to the host's clinical team.
Local councils
Parks and reserves teams gain willing hands for real work, present with a professional throughout, and the council gains visible environmental stewardship and social cohesion in its own suburbs. Council pools and recreation centres can open a door as experience partners, and a community grant can help fund a coordinator's hours.
A note on cost, in your own currency. A pilot runs about $180 to $290 per active participant per week, everything included. That is an order-of-magnitude comparison, not a precise claim, but it sits well below an emergency-motel week, a year of re-imprisonment, or a repeat emergency-department or detox admission. The gap is wide enough that precision is not the point.
It does not duplicate what you already fund. It fills the unfunded rung below your existing contracts, and it can bolt onto a provider you already fund who carries the legal responsibility.
The line that is not negotiable. Purpose Crew is always voluntary and dignity-first. It is never a mandated, sanctioned or work-for-the-dole scheme. You decide what you direct toward it. You never decide whether a participant may walk away.
See it work.
Two things carry a crew: the moment a person is seen, and a way of running the week that fits how people actually live. Here is a look at both.
The moment a person is seen.
At the open or close of a session, the steward names the people who are here and the work they did, in concrete terms. Not a speech about Purpose Crew, a few honest words about the people in the room. "Today we cleared the riverbank by the bridge. That was real work, and the people who use this place will be glad of it." For someone rebuilding a routine and a circle, being named well, out loud, in front of others, is its own kind of repair.
And a rule that protects it. The recognition is about what a person did here, never where they came from. Nobody is singled out as a success story or a rescue. Nobody is named without their say-so. And visitors, funders and photographers never turn the moment into a display. The recognition belongs to the crew, not to an audience.
The tool a chapter uses, when it wants one.
The first chapter needs no software at all. Stamped cards in people's hands and the steward's paper ledger run the whole thing, and the ledger, never the card, is the source of truth, so a token cannot be forged, taken, or pressured out of anyone. When a chapter is ready for it, there is an optional dashboard: people choose their own sessions from their phone and earn tokens toward experiences, a coordinator posts the work and marks attendance, and a printed noticeboard covers anyone without a phone.
A working example, filled with sample data. No chapter is running yet, so every name and code below is illustrative, not a real person.
Browse and book
See this week's sessions and how many places are left, book the ones that suit, and watch tokens add up toward the next experience. Sign in with a demo code: RIMU, KAURI or TOTARA.
Run the week
Post the work, mark who showed up so their token lands, and print a noticeboard for the crew members without a phone. The everyday view a chapter lead would use.
What a funder or an agency actually receives.
Separate from the day-to-day tool, a chapter produces a short, honest report: sessions run, hours contributed, experiences shared, and outcomes told as routine, belonging and readiness rather than a placement count. It shows attendance at around half, because that is normal and planned for, not hidden. It is the evidence artifact, not a highlight reel.
The model, the kits, the brand, and this working dashboard, all built and gifted free.
A founding Auckland host, the two legal confirmations, and the first crew on the ground.
An honest published pilot, so the second chapter can start cheaply and quickly.
A chapter anywhere in New Zealand, started in a weekend on a pack a stranger could follow.
The roadmap is what we are building toward, not a claim about today. Today, the honest headline is that a running pilot is the better part of a year away.
If this could belong where you are, start a conversation.
No cost to take the model, and no obligation in a first conversation. The next chapter walks through how it stays safe, and it is the part most organisations want to read closely.
Get in touchHow it stays safe.
This is the part to read closely. Good intentions are not a legal structure, and this model only works if it is honest about the law it sits inside. Nothing here says "it is fine," and you should be wary of anyone who does. These are named risks with real mitigations, and two of them are hard gates.
- Two hard gates, cleared in writing before a single participant joinsAn employment-law opinion on whether a token counts as a "reward," and an MSD confirmation on whether a token touches a benefit. No opinion, no pilot. That one line is the most important in this whole walk-through.
- A chapter lives inside a host that already carries the dutyThe host organisation is the legally responsible body, runs Police vetting, and holds public liability cover. That is the whole reason a chapter lives inside a trusted host and never stands on its own.
- ACC already covers personal injuryIn New Zealand this is settled, so it is not an open question. The only insurance step is confirming the host's public liability cover reaches off-site work, with one broker quote before launch.
The detail behind each is here for the reader who wants it, and folded away for the one who does not.
The employment-law gate, in detail
Under the Employment Relations Act 2000, a volunteer is someone who neither expects nor receives reward. A token is a reward for time given, which is the exact fact pattern that could pull this toward minimum-wage and employment obligations. New Zealand courts read "reward" broadly: in Salad Bowl v Howe-Thornley (2013) even a free salad counted, though there the worker also expected money and worked for a commercial business, two features absent here. Our design protections are real: taking part is optional and unrostered, nobody loses anything by staying away, tokens never convert to cash, there is no obligation to finish a task, and no task displaces paid work. But design is not a ruling, so a written opinion scoped to section 6 of the Act is obtained, pro bono, before recruiting a single participant.
The benefit gate, in detail
If a token were counted as income under the Social Security Act 2018, taking part could cost a participant money, a cruel outcome. The token is designed to sit outside income: non-cash, non-convertible, non-transferable, spendable only on experiences the chapter organises. Designed is not confirmed. The route is that the host's MSD contracts manager puts the question to MSD national office in writing, with a plain description and a clear deadline, because local offices do not issue binding rulings. If no written confirmation comes, the fallback is a legal opinion on the income definitions plus full, plain-language disclosure to every participant before they join. Never assume, and never let the first answer arrive as a debt letter.
Health, safety and vetting, in detail
The Health and Safety at Work Act 2015 applies to volunteers, so the duties are real and need a named owner: the host, as the person conducting the undertaking, written into the hosting agreement along with its access to Police vetting. Practical controls: the low-hazard task menu, a short hazard induction each session, at least two vetted adults present and never one-on-one out of sight, an incident log reviewed within 48 hours, and first aid cover with a kit that travels with the crew.
Non-displacement is a signed line, not a hopeful rule. A task qualifies only if it is group-based, low-hazard, has a visible result, and displaces no paid work. That last test is a written confirmation, with a twelve-month look-back, signed by every work host: nobody is, or recently was, paid to do this. It is the single control that answers the "cheap labour" worry before it is raised.
What we ask, and what we never ask.
What we ask at intake
A name to go by. How to reach you, if you want to be reached. Anything that keeps you safe in a group. One cup of tea, twenty minutes, no forms.
What we never ask
Benefit status. Criminal history. Immigration status. Proof of address. How you came to be here. None of it is any of our business, and none of it decides whether you belong in the crew.
This is why the benefit gate matters, and it is worth being plain about it: the question is not really about protecting the host. It is about making sure that showing up can never quietly cost a participant money. That is why it is a gate, and not a footnote.
What we have not yet answered. An honest list, each paired with the step that settles it.
- Whether a token counts as "reward" in law.
- The written employment-law opinion, before anyone joins.
- Whether MSD treats the token as income.
- The written question through the host's contracts manager, with the disclosure fallback.
- Whether attendance holds near half.
- Over-recruit, record it honestly from week one, and publish the real figure with the results.
- Whether a steady plateau is a success or a limit.
- Reviewed with each participant at six months, against their own stated goals, not a placement count.
- Whether a model designed by a Pākehā builder fits communities where Māori lead.
- Iwi and kaupapa Māori partnership at governance level from the first conversation, and a te reo name gifted through consultation.
One boundary is firm alongside all this: Purpose Crew stays non-clinical. It never assesses, treats or counsels, and any clinical concern goes straight back to the host's clinical team. That is another reason it lives inside an established host.
Two of these run on clocks nobody controls. The employment-law opinion and the MSD letter each take weeks, with no service standard, so they are started on day one, before anything participant-facing happens. When you reach How to start, you will see them as the very first actions.
Council or Crown agency? Here is what changes
The version you can put in a paper
Everything above, as a risk-and-mitigation register you can lift straight into an internal case. Non-displacement is the headline control, because it is the one your probity people will look for first.
| The risk | How it is held |
|---|---|
| Token counts as a "reward" in law | Scoped written employment-law opinion on ERA section 6, obtained before any participant joins. No opinion, no pilot. |
| Token reduces a benefit | Written MSD confirmation via the host's contracts manager, with a legal-opinion-plus-disclosure fallback. |
| Displacing paid work | Written non-displacement rule, twelve-month look-back, signed by every work host. The headline control. |
| Health and safety duty | Host is the named PCBU, runs Police vetting, low-hazard tasks, two vetted adults per session, first aid. ACC covers injury. |
| Coercion or exploitation optic | Voluntary and unrostered, tokens never cash, participant co-design, whole model published openly. |
If your agency already contracts a provider who could host, the path is shorter still: their vetting, insurance and legal responsibility already exist, so a chapter can bolt on rather than start from scratch.
What a pilot takes.
Here is the honest shape, with every figure marked as a planning estimate to re-cost against local rates. A pilot is one chapter, six months. Recruit 16 to 18 people, expect around 12 active in any week, because attendance near half is normal for people whose lives are unstable, and the budget is built for that, not for perfection. The all-in estimate is $55,000 to $90,000 for six months, roughly $180 to $290 per active participant per week, everything included. A chapter that builds a strong base of donated places, tools and seconded hours can roughly halve the cash it needs to raise.
Three money lines, stated to everyone from the start.
- No cash to participants, ever. Tokens never convert to cash, are named to the person, and are spent on shared experiences the chapter organises. This is what keeps the model dignified and outside the wage and benefit traps.
- No founder salary from chapter funds. The money a chapter raises pays for that chapter's delivery: facilitation, transport, materials, the experiences. It never pays the builder.
- Shared systems funded separately, so every local dollar stays local. The model and the website are gifted free. Keeping the shared systems current is a paid service Filipp provides as a retained systems advisor, funded from the AI and I layer, disclosed openly from the first conversation, and never drawn from a chapter's budget.
The money is layered so no single funder carries the pilot, and every grant is applied for by the host organisation, the registered charity with the contracts, payroll and audit, so the money lands where the work happens. The biggest line of all is in-kind: donated seats, venue space, tools, a van on loan, seconded hours. A good tactic is to secure one anchor pledge before asking anyone to commit to the pivotal hire, so the funding circle is broken rather than stalled.
That pivotal hire is a founding coordinator, funded before the role is offered, and the plan for lasting is to hand it over to the host's own staff. Real budget lines sit alongside it too: external supervision for the facilitators themselves, and three paid co-design workshops with people who have lived experience, before launch, so they shape the tasks, the rewards and the rules. And hosting is ring-fenced and exitable, with a clear indemnity-and-exit clause, so a host can step back without the model collapsing.
The honest timeline, and the low-barrier start, together. Launch-ready is the better part of a year out, because the host search and the two legal clocks are slow and have no shortcut. Say that plainly, so nobody under-promises a start date. And in the same breath: you can begin on Monday with a landing page, a spreadsheet and a paper ledger. No software, no budget, nothing to buy.
The launch gates, as one checklist.
- A hosting agreement signed, naming the host as the legally responsible body and committing it to run Police vetting.
- Both legal gates cleared in writing: the employment-law opinion, and the MSD confirmation or its disclosure fallback.
- Three facilitators funded in cash or covered by signed secondment agreements.
- At least four experience partners on dated, signed three-month commitments.
- Public liability cover confirmed by a broker quote. Any gate open, the start date moves. The gates do not.
One thing that can help: once a host holds charity status, it can unlock a Google Ad Grant of up to about USD $10,000 a month in free search advertising, which also happens to be the kind of shared system the advisory role keeps current.
Council or Crown agency? Here is what changes
The open door
This is where the open door is most real. The funding stack above is built for community trusts, because that is who funds most community work. A Crown agency does not sit in that stack the same way. You are not another grant to apply for. You already hold channels a community trust does not, and you are the one who knows which of them could fit.
How an agency actually engages, plainly. Crown money does not move by "direction," it moves through defined routes. A few that fit: a low-value direct engagement below the open-tender threshold; a variation on a contract you already hold with a provider who could host; an innovation or evidence fund; or a co-design memorandum first, with no money attached, to shape it together. The model is gifted, there is no licence and no lock-in, and the counterparty is always the host charity, not the builder.
What your agency gets back. Not access to people. An evaluated proof of concept you can point to, the data and evidence from it which you own, and a low-threshold rung you can refer people to, the step below your existing contracts that does not exist today. The same three money lines hold in procurement-probity terms: no cash to participants, no founder salary from delivery funds, and the systems layer funded and disclosed separately so every local dollar stays local.
A shape that fits a budget year. If it helps, this can come in two phases: a small scoping engagement this financial year to shape and de-risk it, and a pilot the next. And the line that does not move: you decide what you direct toward it, you never decide whether a participant may walk away.
Where you fit, and how to start.
Almost everything hinges on one thing: a host. The host is the keystone that makes every other door possible, the people, the partners, the funding, because it holds the trust, the premises and the legal responsibility. Get the host right and the rest follows. Lose the host and the whole thing falls, which is exactly what happened to the schemes this one learned from.
The one early question. Everything else can wait, but this cannot: would your organisation, in principle, be willing to be the named legally-responsible body and run Police vetting for a ring-fenced pilot? A single early conversation can settle it, and because it reshapes everything, it is worth settling first.
Ways to be part of it.
Host a chapter
Be the home a crew runs from, at your own pace, with our support throughout. This is the one that makes everything else possible.
Refer someone you support
When a person you work with needs routine, purpose and company, a crew can be that next gentle step. Nobody is a failed referral, and someone turned away today is welcomed another day.
Offer the work
A garden, a clean-up, a sorting job: one supervised session a month, facilitator present throughout, with a short written non-displacement confirmation we draft for you.
Offer an experience
A cinema, a pool, a barber, a bowling alley: a few off-peak places a month for a supervised group, booked ahead, no cash, and a warm named thank-you.
Fund a chapter
Take a place in the funding stack behind a host, as an anchor grant, an operating backbone, or in-kind toward the experiences and transport.
Run it in your city, or help it spread
Take the whole model and start a chapter where you are, or simply point us to a community, or an organisation, that this could belong to.
What to do first, on Monday.
- List the eight to twelve organisations near you that could be a host.
- Ring the warmest one and put the single early question above.
- Open the employment-law opinion, pro bono, through a Community Law Centre or a firm's community programme.
- Draft the MSD question for the host's contracts manager to send.
- Start two lists: work and experience partners, and funders.
The first two cost nothing but a phone call. The middle two start the slow clocks running quietly in the background. None of it needs a budget, or a single line of software.
What "request the free pack" actually gives you.
The whole model is gifted free, to take, adapt, and run under your own name. Nothing in it is proprietary.
Two doors, and the choice is yours.
You can take the whole model and run it yourself, or start a conversation and we will walk it through with you. There is no cost either way, and no obligation in a first talk. If it could belong where you are, we would love to hear from you.
Get in touch Read the full prospectusCouncil or Crown agency? Here is what changes
A lighter first step, and something to forward
The early question for an agency is gentler than for a host, because you may not be the legal home yourselves. It is simply: is there an agency home, and a referral relationship, we could start from? That is enough to begin.
And rather than the host pack, which is shaped for the organisation that runs a chapter, there is a one-page business case shaped for an internal conversation: the problem, the evidence, the cost, the risks, the specific ask, and what your agency gets. Something you can forward to a manager or a committee without rewriting it.
Open the one-page business case
The lighter first step is a short, tailored fit-note for your agency, or a scoping conversation, a hui or a memorandum. Never anything that reads as a funding commitment. The door stays open, and the pace is yours.
A last reminder: this page, and every way to reach us on it, is for hosts, partners, funders and agencies. Places on a crew are never arranged here. Anyone who needs support today is warmly pointed to a live service.
