Give People a Job, Not Just a Seat: Roles Inside Social Prescribing Groups

A link worker refers someone to a weekly walking group, a craft circle, a men’s shed. They show up for three sessions and stop. Ask why, and the answer is rarely “I didn’t like it.” It’s closer to “I didn’t know what I was for.” Attendance without a role is a fragile thing to build retention on, and most social prescribing groups are designed around attendance alone.

The fix is not more warmth. It’s structure. Give people something to do inside the group, not just somewhere to be.

Why “just showing up” fails by week four

Qualitative research on social prescribing has found that participants describe the benefit as something bigger than contact itself — restored participation, a sense of contributing, purpose. Structured, purposeful activity appears to outperform loose social contact as a mechanism for benefit. That’s a specific finding, not a general one: it says the activity design matters, not just whether the room is friendly.

Call the collapse point the week-four problem. It shows up across programme types that rely on passive membership: novelty carries the first few weeks, then attendance depends entirely on whether the person feels like an audience member or a participant. A group with no roles has no mechanism for someone to become a participant. Everyone is equally welcome and equally replaceable, which is exactly the condition under which people quietly stop coming.

This maps onto older thinking about informal social life. Ray Oldenburg’s account of the third place emphasizes that these settings work because they have regulars — people with a recognized position, not just visitors. Robert Putnam’s distinction between bonding and bridging social capital points the same direction: groups that generate durable ties tend to have some internal division of labor, not just shared presence.

What a role actually looks like

A role in this context is small, visible, and repeatable. It doesn’t need a title or training. Examples that fit inside an existing social prescribing group without redesigning the whole programme:

1. Greeter — arrives ten minutes early, welcomes new referrals, points them to the coffee.
2. Setup/pack-down — owns the physical space each week: chairs, materials, kettle.
3. Continuity keeper — remembers who was there last week, notices who wasn’t, mentions it.
4. Convener’s second — helps run the session when the facilitator is away, so the group doesn’t fold if staff are absent.
5. Recruiter — brings a friend, or tells the link worker who else might benefit.

None of these require competence beyond what any regular attender already has. That’s the point — the role should be achievable in week two, not reserved for people who’ve been coming for a year.

There is a genuine evidence base for the underlying mechanism, if not for this exact tactic. A randomised controlled trial of volunteering among lonely older adults in Hong Kong — one of the few RCTs in this literature rather than an uncontrolled evaluation — tested prosocial engagement against a control condition and is one of the only pieces of trial-grade evidence on whether doing something for others changes loneliness outcomes, as opposed to simply receiving social contact. That trial is about volunteering specifically, not in-group roles, and it should not be stretched further than that. But it’s the strongest available signal that agency and contribution, not exposure, are doing the work.

What this means in practice: before a group’s third session, assign at least one small, named job to at least a third of attendees. Rotate it. Don’t wait for people to volunteer — most won’t, and the ones who would have anyway are not the retention problem you’re trying to solve.

Evidence status

Claim	Evidence status
Social prescribing groups suffer high attrition after initial sessions	Widely reported in programme literature; systematic reviews note heterogeneity and weak retention data across studies
Structured, purposeful activity produces more benefit than contact alone	Supported by qualitative meta-synthesis of participant accounts
Assigning specific in-group roles improves retention	Plausible mechanism, not directly tested; inferred from third-place and social capital theory plus adjacent volunteering trial
Volunteering/prosocial engagement reduces loneliness in older adults	One RCT (Hong Kong), not yet replicated
Social prescribing reduces loneliness at all	Reviews report positive outcomes across included studies, but a 2025 protocol notes only one peer-reviewed RCT exists for older adults specifically — the field runs mostly on uncontrolled evaluation

That last row matters more than the others. Social prescribing as a field has enthusiastic qualitative support and thin experimental support. A 2025 systematic review protocol makes the point plainly: despite growing adoption, the effectiveness of social prescribing for older adults remains unclear, and there is essentially one trial to point to. Programme managers should hold the role-assignment idea in this piece at the same evidentiary weight — a reasonable, theory-backed operational tweak, not a proven intervention.

What to track if you try this

Don’t just ask whether people liked having a role. Track:

• Session-to-session attendance rate for role-holders versus non-role-holders over eight weeks.
• Whether role-holders are the ones who eventually recruit others in, which is the mechanism you actually want.
• Drop-off specifically at weeks three to five, since that’s where the effect (if real) should show up first.

If role-holders don’t outlast non-role-holders by week six or so, the tactic isn’t working for that group and you should stop insisting on it rather than assuming the roles were just badly designed.

What this does not solve

This is a retention tactic for people who are already inside a group. It does nothing for the referral gap — the fact that social prescribing, like most programme design, disproportionately reaches people willing and able to take a referral in the first place, and misses the isolated person who never gets to a link worker at all. It also does nothing to address a structural third-place shortage: recent geographic research has found third-place closures concentrated in socially vulnerable and rural areas between 2019 and 2021, meaning some communities don’t have the physical setting in which any of this — roles, regulars, or otherwise — could happen. Roles inside a group are a low-cost fix for a real design flaw. They are not a substitute for having somewhere for the group to exist.