What Funders Are Actually Buying: Service Delivery or Community Membership

A bid lands on your desk asking for £40,000 to run a weekly walking group for twelve months. The logic model has referral numbers, attendance targets, and a UCLA Loneliness Scale pre/post. It looks fundable. It also looks exactly like a bid for a twelve-week befriending pilot, a six-month peer support group, and a permanent community choir. These are not the same thing, and funding them as if they were is a large part of why so many connection programmes collapse the moment the grant does.

The distinction that matters is whether you are funding a service — something delivered to people, with a start, an end, and a throughput target — or a community — something people belong to, that has no natural end and whose value increases the less it depends on paid staff. Both are legitimate things to fund. They need different money, different staffing, different outcome measures, and a different theory of what “success” looks like at year three.

Two different objects, dressed the same in a bid

Robert Putnam’s distinction between bonding and bridging social capital is useful here, even though it was built for a different argument. A service produces bridging contact: it connects a person to an activity, a professional, or a moment of company, and its unit of success is the interaction. A community produces bonding capital over time: it produces a set of people who recognise each other, show up for each other outside the programme, and eventually don’t need a facilitator to convene them. Ray Oldenburg’s account of the third place describes the endpoint — a bar, café, or club that survives because it belongs to no one and everyone, not because an organisation keeps commissioning it. Eric Klinenberg’s argument about social infrastructure makes the same point about physical space: the thing that outlasts a grant cycle is rarely the programme, it’s the place and the relationships the programme happened to seed there.

Most funding instruments are built for services. They want a start date, an end date, a discharge criterion, and a number that moved. Systematic reviews of social prescribing report exactly this kind of outcome — improved self-esteem and confidence over a defined referral period — but note limited trial evidence and wide variation across programmes, which is a polite way of saying nobody has agreed what a successful referral episode looks like once you go looking for it. A qualitative synthesis of social prescribing found something more interesting underneath the averages: participants describe benefit that extends past social contact into restored purpose and participation — the language of belonging, not of throughput.

That is a community outcome showing up inside a service contract. It is also usually the first thing an evaluation drops, because it doesn’t fit the logic model the funder asked for.

Where the money actually needs to go

Dimension	Service	Community
What you’re paying for	Facilitation, referral pathway, defined session count	Space, low-key stewarding, continuity beyond staff turnover
Natural exit	Discharge after a set number of sessions	None — the question is sustainability, not exit
Right outcome measure	Change score over a fixed window (e.g. loneliness scale pre/post)	Retention, self-organising activity, member-initiated events
Staffing risk	Facilitator leaves, service ends on schedule	Facilitator leaves, community either survives or dies depending on whether roles were ever shared out
Funder’s real question	Did the intervention work?	Does this still exist without us?

The failure mode worth naming is funding a community as if it were a service. It shows up as a grant that pays for twelve structured sessions, ending with a “graduation,” for a group that was actually becoming a self-sustaining walking club. The funder closes the grant on schedule, satisfied the logic model held. The group, cut off from the room booking and the WhatsApp admin nobody transferred, dissolves within a month. The outcome data still says the intervention “worked” — because it was measured at week twelve, not week sixteen.

The reverse failure is rarer but real: funding a service as if it were a community, meaning underspecifying facilitation and staffing on the theory that “it’ll find its own feet,” when the target group — people in acute crisis, recent bereavement, severe isolation — needed structure, not a room and good intentions. The Foundation for Social Connection’s peer network and the Campaign to End Loneliness’s own work on measurement exist partly because commissioners kept conflating these two asks in a single grant line.

What this means in practice: decide, before you write the bid or review it, whether you are buying an outcome that happens once (a service) or a set of relationships that need to keep happening without you (a community). If it’s the latter, budget for a transition period where facilitation deliberately steps back and roles get handed to members — and fund that transition explicitly, not as an afterthought inside the final quarter.

Evidence status

Claim	Status
Structured, purposeful group activity outperforms unstructured contact alone	Reasonably supported — qualitative synthesis evidence
Social prescribing improves self-esteem and confidence	Supported but with limited, heterogeneous trials
Physical third places are declining and that decline tracks with connection loss	Documented at population level (library visits, bowling centres, social clubs), not causally proven for any single programme
A named transition period improves community survival past grant end	Plausible from theory (Oldenburg, Klinenberg); not directly tested in the social prescribing literature reviewed here

What this does not solve

None of this tells you which model — service or community — is right for a given population, and the underlying evidence base for social prescribing is thin enough that a 2025 systematic review protocol still found only one peer-reviewed randomised controlled trial in the entire field for older adults. Nor does it solve the reach problem: distinguishing service from community funding changes what happens to the people who already found their way into a group. It does nothing for the people who never got referred at all, and the decline in the physical spaces — libraries, bowling centres, bars — where communities used to form on their own means fewer of those informal routes exist to begin with.