The First Four Weeks: A Sequencing Guide for New Connection Programmes

Programmes rarely fail at launch. They fail in week four, when the novelty that got people through the door has worn off and nothing structural has replaced it. If you are designing or restarting a connection programme — a walking group, a men’s shed, a social prescribing referral pathway, a workplace peer network — the first month determines whether it survives contact with ordinary life. Below is a sequence, not a philosophy.

Week one: recruit for the wrong reason, on purpose

Most people join a connection activity for a reason that has nothing to do with connection — a doctor told them to, a friend asked, they wanted to learn pottery. That is fine. Do not try to correct this in week one by talking about loneliness or wellbeing. AARP’s 2018 survey of adults 45 and older found that the strongest predictors of loneliness were network size and diversity, and physical isolation — not attitude. People do not need convincing that connection matters. They need a reason to show up that survives a bad week.

Practical task for week one: confirm attendance by a specific, low-effort mechanism — a text, not an email; a named contact, not a general inbox. Get a second commitment before the first session ends. “See you Thursday” beats “hope to see you again.”

Week two: this is where structure earns its keep

A qualitative meta-synthesis of social prescribing found that participants describe the benefit of these programmes as extending beyond social contact itself, toward restored participation and purpose — and that structured, purposeful activity appears to work better than unstructured contact alone. This is the single most practically important finding in the literature for sequencing, because it tells you what week two is for: giving people a role, not just a seat.

That means by week two, someone new should have a task — setting up chairs, leading a section of the walk, bringing something next time. Not because busywork is virtuous, but because a role is what turns a session someone can skip into a session someone is needed at.

What this means in practice: if your week-two session is structurally identical to week one — same format, no roles assigned, no reason to come back other than “it was nice” — you are relying on goodwill to do the job that structure should be doing. Goodwill runs out around week four.

Week three: address the free-rider problem in befriending models

If your programme is a one-to-one befriending model rather than a group activity, week three is where you should be watching hardest. A 2025 randomised controlled trial of befriending in residential aged care found it did reduce loneliness scores meaningfully at eight and sixteen weeks. But a 2026 trial testing telephone-delivered behavioural activation and mindfulness against a befriending control, in a large sample of older adults living in poverty and digitally excluded, found the structured intervention beat straightforward befriending. Befriending works. It is simply not the strongest tool in the box when compared head-to-head with something more directive.

The implication for week three: if you are running a befriending pairing, add a task to the conversation — a shared small goal, a piece of information to report back, a decision to make together — rather than leaving the pair to generate their own structure from nothing. Pure company is fragile at week three in a way that shared purpose is not.

Week four: the drop-off point, and why it happens here specifically

Call this the week-four problem: attendance holds through weeks one to three on the strength of initial commitment and curiosity, then falls sharply once the activity has become familiar enough to skip without guilt. It is the point at which a participant asks, consciously or not, “what happens if I don’t go this week?” — and if the honest answer is “nothing,” many won’t.

The evidence on social prescribing’s effect on service use is instructive here. A systematic review of loneliness-focused social prescribing found that of nine included studies, three reported reductions in downstream health or social service use — not zero, but not universal, and the effect is concentrated in programmes with clearer structure and follow-through, not simply referral. Referral without a structural answer to “what happens if I don’t go” is where a lot of social prescribing money leaks out.

The fix at week four is not more enthusiasm. It is making absence visible and low-stakes to repair: a follow-up contact within 48 hours of a missed session, framed as logistics rather than concern (“we moved the room, see you next week”) rather than as a wellness check, which can feel surveillant.

Evidence status of what actually helps at each stage

Claim	Evidence status
Structured, purposeful activity outperforms unstructured contact	Reasonably supported — qualitative synthesis across multiple social prescribing studies
Befriending reduces loneliness on its own	Supported by at least one RCT, but outperformed by structured alternatives in head-to-head testing
Social prescribing reduces downstream service use	Weak — only a minority of reviewed studies found this, and heterogeneity across programmes is high
A fixed physical venue improves retention	Plausible from social infrastructure theory (Oldenburg’s third places, Klinenberg’s social infrastructure argument) but not tested against alternatives in controlled trials
Follow-up contact after a missed session reduces drop-off	No dedicated trial evidence; inferred from the general pattern that structure and role assignment outperform contact alone

What this does not solve

This sequence is built for people who have already shown up once. It says nothing about reach — the much larger and harder problem of people who never hear about the programme, don’t trust the referral pathway, or face a transport or cost barrier that no amount of week-four follow-up will fix. Social prescribing evaluations consistently note this limitation: they measure what happens to people already inside the system, not who never entered it. A perfectly sequenced four weeks will retain the people you recruited. It will not recruit the people you missed.