How to Tell If You're Measuring a Service or a Community

A weekly walking group and a twelve-week befriending programme can produce identical satisfaction scores and mean completely different things. One might be building something that outlives its funding cycle. The other might be a well-run service that evaporates the day the sessions stop. Most evaluation frameworks cannot tell the difference, because they were built to measure delivery, not durability.

This matters because commissioners increasingly want to fund “community,” not just “sessions attended,” but the metrics you already have — attendance, satisfaction, referral completion — are service metrics. They tell you whether you ran the thing you said you’d run. They tell you almost nothing about whether participants formed bonds that will hold once you withdraw.

Step 1: Separate what you deliver from what forms between participants

A service is something an organisation provides to a person. A community is something that exists between people, independent of the organisation providing it. Robert Putnam’s distinction between bonding and bridging social capital is useful here, but the more basic test is simpler: if your organisation disappeared tomorrow, would the connections your programme created survive?

For most social prescribing referrals, walk-in drop-ins, and time-limited group programmes, the honest answer is no. A qualitative meta-synthesis of social prescribing found that participants describe benefit extending beyond the contact itself — restored purpose, a reason to leave the house — but that benefit is frequently tied to the structure of the programme, not to a self-sustaining peer network. That is not a failure. It is a service doing what services do. The failure is measuring it as though it were something else.

Step 2: Audit your current metrics against this table

What you’re measuring	What it actually tells you	Service or community signal
Attendance rate	Whether people show up while the programme runs	Service
Satisfaction score	Whether people liked what was delivered	Service
Referral completion	Whether the pathway worked administratively	Service
Contact exchanged between participants outside sessions	Whether a bond formed independent of staff	Community
Attendance six months after the programme formally ends	Whether the group persisted without organisational scaffolding	Community
Participants organising their own follow-up activity	Whether ownership transferred from staff to members	Community
Someone missed and someone else noticed	Whether the group has developed mutual accountability	Community

If every row you can currently fill in sits in the left column, you are running — and measuring — a service. That may be exactly the right thing to run. But say so to funders, rather than reporting service metrics as if they were evidence of community.

Step 3: Build in a measurement point after you leave

The single most diagnostic thing you can do is measure after the programme ends, not just during it. Most social prescribing evaluations stop at discharge or shortly after, which is also where most service-only benefit shows up and where community benefit, if it exists, has not yet had time to prove itself. A systematic review of social prescribing and loneliness found that all nine included studies reported positive individual impacts during the intervention period; it is much harder to find comparable follow-up data on what happens after.

Set a check-in at three and six months post-programme, asking not “how are you feeling” but three narrower questions:

1. Are you still in contact with anyone from the group?
2. Has that contact continued without any involvement from staff or the organisation?
3. Has anything new grown out of it — a separate meet-up, a shared activity, a different group entirely?

A “no” to all three at six months is not proof the programme failed. It is proof the programme was a service, and should be evaluated, funded, and staffed as one.

Step 4: Look for the mechanisms that actually produce community, not just contact

Ray Oldenburg’s account of third places argues that informal recurring spaces — cafés, barbershops, pubs — generate community precisely because they are unstructured, repeatedly visited, and low-stakes. Eric Klinenberg’s work on social infrastructure makes a related point about physical spaces: libraries and parks shape rates of contact because people return to them on their own initiative, not because a facilitator schedules them to. Programmes that mimic this — a standing weekly slot in the same physical location, minimal facilitation, permission to skip and return — are more likely to generate self-sustaining ties than programmes built around a fixed curriculum with a defined endpoint.

The randomised trial of volunteering among lonely older adults in Hong Kong is instructive on this point in a different way: it is one of the few controlled tests of a connection intervention rather than an uncontrolled programme evaluation, and it tested prosocial engagement — giving participants a role, not just a seat — against a passive control. Roles that persist beyond the session are more likely to generate the kind of mutual accountability that a service alone does not.

What this means in practice: if your logic model has a defined end date, a fixed curriculum, and staff present at every session, you are designing a service. That’s a legitimate design choice, but measure it with service metrics — completion, satisfaction, referral outcomes — and don’t claim community impact you haven’t tested for. If you want community, redesign for continuity beyond your presence: a standing time and place, a handover of ownership to participants, and a follow-up measurement point after you’ve left the room.

The week-twelve handover problem

Programmes that do intend to build lasting community fail at a predictable point: the moment staff withdraw. Call it the week-twelve handover problem. Facilitators run twelve tight, well-attended sessions, then step back, assuming the group will now run itself. Usually it doesn’t, because nobody was given the tasks that make groups self-sustaining — who books the room, who messages people who missed a week, who decides what happens next. Community capacity has to be built into the programme deliberately, from an early session, not assumed to appear once staff step back.

What this does not solve

None of this fixes the underlying evidence problem in the field: most social prescribing research is heterogeneous, short-term, and short on the kind of controlled comparison the Hong Kong volunteering trial represents. A six-month follow-up call tells you whether contact persisted, not whether it meaningfully reduced loneliness or improved health — those require the kind of outcome measurement most local programmes are not resourced to run. And this framework says nothing about reach: it helps you measure what happens to the people who joined, not the far larger number who never had a programme, a service, or a community within reach to join in the first place.