How to Split Roles Inside a Group So It Outlives the Grant

Every funder asks the same question near the end of a grant application: what happens when the money stops. Most applicants answer with a plan to find more money. That is not a sustainability plan. It is a hope.

A better answer is structural: design the group so that no single funded role is load-bearing. This means deciding, before the first session, who does what — and building enough redundancy that the departure of a paid facilitator is an inconvenience rather than a death sentence. This is a design problem you solve once, at setup, not a staffing problem you solve every time someone leaves.

The founder cliff

Call it the founder cliff: a group runs well for eighteen months under a funded coordinator, then the grant ends, the coordinator’s contract ends with it, and attendance drops by half within two sessions and to zero within two months. This happens because one person was quietly performing four or five distinct functions — booking the room, remembering names, deciding the agenda, managing the WhatsApp group, closing the door at the end — and nobody else knew which functions mattered or how to pick them up.

Robert Putnam’s account of declining American civic participation in “Bowling Alone” is partly an account of this: organisations that depended on one or two indispensable organisers didn’t survive those organisers’ retirement, illness, or boredom. The lesson generalises to funded programmes. A group’s resilience is a function of how distributed its labour is, not how committed its founder is.

Five roles, named separately

Most groups that run without paid staff — the ones Meetup was originally built around, and the kind Ray Oldenburg described in his account of the third place — divide labour across recognisable functions even when nobody has written them down. The fix is to write them down, on purpose, before you need them.

1. The convenor. Decides when and where the group meets and holds the calendar. This is administrative, not charismatic. It can be done by someone with thirty minutes a week and a shared document.

2. The host. Physically opens the space, sets up chairs, makes the room legible to a first-time attendee. Eric Klinenberg’s argument in “Palaces for the People” is that social infrastructure works because it lowers the cost of showing up — someone has to actually do that lowering, session by session, and it is a distinct job from deciding the room exists.

3. The connector. Notices who is new, makes an introduction, remembers a detail from last time and raises it this time. This is the role most people assume is “just what a friendly person does.” It is not automatic, and if it is not assigned it usually does not happen consistently — which matters, because network size and diversity are the strongest predictors of lower loneliness in the 2018 AARP Foundation survey of adults 45 and older, and a connector is the mechanism by which a group actually expands someone’s network rather than just occupying an hour of their week.

4. The keeper of continuity. Holds the group’s memory across absences — who missed last time, who mentioned a hospital appointment, what the group decided about format last month. Without this role, every session restarts from zero and long-term members drift because the group has no memory of them.

5. The closer. Ends the session on time, on a clear note, and confirms the next date out loud before people leave. This sounds trivial. It is the single most commonly dropped role when a facilitator is stretched, and its absence is what makes sessions feel like they are winding down rather than continuing.

None of these five roles requires a salary. All of them require someone specific to own them, by name, in a document the group can see.

How to assign them without recreating the same bottleneck

1. Write the five roles down before recruiting attendees, not after the group has formed its own informal hierarchy. Retrofitting roles onto an established group is harder because people have already decided, without discussion, who “is” the leader.

2. Assign at least two people to each role from the start, even if one is clearly the primary. A role with a single owner is a role with a founder cliff built into it.

3. Put the paid coordinator in the convenor and continuity roles, not the host and closer roles. Administrative and memory functions are the ones a funder’s staff time is actually suited to. Host and closer are the roles most naturally handed to members, because they require presence and consistency more than training.

4. Review role coverage at the funding midpoint, not the end. If the connector role is still resting entirely on the paid facilitator six months before the grant ends, that is the point to fix it — not the final month, when there is no runway to test whether the fix works.

5. Tell the funder about the role structure, not just the activity. A line in a report that says “five distinct roles are now held by seven volunteer members, with two roles held jointly by staff and volunteers” is a more credible sustainability claim than “participants report high satisfaction,” because it describes a mechanism rather than a mood.

What the evidence actually supports here

Claim	Evidence status
Structured, purposeful group activity produces more benefit than unstructured social contact	Reasonably well supported — a 2022 qualitative meta-synthesis found participants described benefit tied to restored purpose and participation, not contact alone
Network size and diversity predict lower loneliness	Well supported by the 2018 AARP Foundation survey of adults 45 and older
Distributing organisational labour prevents the “founder cliff” specifically in funded groups	Plausible from civic-participation literature and social infrastructure theory, but not directly tested — no controlled study measures group survival by role-distribution design
Social prescribing groups generally improve confidence and self-esteem	Supported by a 2021 systematic review, though the same review flags limited trial evidence and high heterogeneity across programmes

The middle row is the honest gap. Nobody has run a trial comparing groups with named, distributed roles against groups with a single informal leader and measured which survives longer post-funding. The claim above rests on organisational theory and pattern recognition from civic life, not on a study built for this specific question. Say that plainly if a funder asks for the citation, rather than dressing up the AARP or Putnam findings as something they are not.

What this means in practice: before you write “sustainability plan” in a grant application, write the five roles on a whiteboard with names next to them — real names, not “TBC.” If any role has only one name against it, that is the sentence a reviewer should be able to circle and ask you about, and the sentence you should be able to fix before submission.

What this does not solve

Role design helps a group that already has people in the room keep meeting once the funding ends. It does nothing for the people who never joined in the first place — the ones the group’s outreach never reached, who are often the most isolated and the least likely to walk into an unfamiliar room regardless of how well the roles inside it are distributed. Structure fixes retention. It is not a recruitment strategy, and a funder’s report should not conflate the two.