The First Four Weeks: Sequencing a New Connection Programme

Most connection programmes lose the majority of their participants before they have a chance to work. Not at week one, when curiosity carries people, and not at week twelve, when the programme is either clearly working or clearly cancelled. They lose them at week four — the point where novelty has worn off, the facilitator has stopped introducing themselves every session, and nothing has yet replaced the initial reason for showing up. Call it the week four problem. It is a scheduling and design failure, not a motivation failure in participants, and it is fixable with sequencing.

This is not about programme content — what activity you run, what curriculum you follow. It is about what order things happen in during the first month, because order determines whether a group becomes a place people return to or a series of unconnected sessions that happen to share a room.

Why order matters more than content here

Ray Oldenburg’s account of the third place describes something that functions as regular, low-stakes, and self-sustaining — a place people return to without being asked. Robert Putnam’s account of civic decline describes the opposite: associations that require increasing effort to sustain participation in, until people stop. A new programme starts firmly in Putnam’s territory. Every session requires an active choice to show up. The job of the first four weeks is to move it toward Oldenburg’s territory, where showing up is a habit rather than a decision. That shift does not happen by accident, and it does not happen evenly across four weeks — it happens in a specific sequence.

Week 1: reduce the entry cost, not the content

The first session should ask less of participants than any session after it. This sounds obvious and is routinely ignored. Programmes over-design week one — icebreakers, forms, introductions, a full explanation of the twelve weeks ahead — because organisers are anxious to prove value immediately. The effect is to make the highest-friction session the first one a participant experiences, at the moment they are least invested and most likely to decide the whole thing isn’t for them.

Do the logistics before week one, not during it: intake forms, expectations, transport information. Week one itself should contain one clear, low-stakes shared activity and nothing else. The AARP’s 2018 survey of adults 45 and older found that people who had spoken to a neighbour, even briefly, were far less likely to report loneliness than those who never had — 33% against 61%. The mechanism there is not depth. It’s frequency of low-stakes contact. Week one should manufacture one instance of that, not attempt intimacy.

Week 2: name the return, don’t assume it

By the second session, attendance will already have dropped — often 10 to 20% is normal even for programmes that go on to succeed. The mistake is treating this as a bad sign rather than a design point. Week two is where you should explicitly acknowledge the group as continuing, not starting over: reference something specific from week one, and ask a returning participant, not the facilitator, to summarise it for anyone new. This does two things. It signals that the group has a memory, which is the first marker of a place rather than an event, and it gives a returning participant a role, which correlates with why people stay in structured group activity in the first place. A qualitative synthesis of social prescribing participants found that the benefit people described was rarely social contact alone — it was restored participation and purpose, doing something that mattered within the group, not just being present in it.

Week 3: introduce a task that survives the facilitator

This is the week to hand something over. A rotating role, a small responsibility, a decision the group makes rather than one the organiser makes for them. Programmes that remain entirely facilitator-run tend to plateau in attendance around week three or four, because the group has no structure independent of the person running it — if that person is not there, or when the programme eventually ends, nothing continues. Eric Klinenberg’s account of social infrastructure makes the related point at a larger scale: places that sustain contact do so because the structure itself, not any single organiser, holds people there. A four-week programme can’t build a building, but it can build a rotating task.

Week 4: check what’s actually happening between sessions

The fourth session is the point to ask, directly, whether any contact is happening outside the programme — a message, a coffee, a lift shared. This is the actual outcome you are testing for, and week four is early enough to still adjust format if the answer is no. A systematic review of social prescribing and loneliness found that most reported benefits came from initiatives with follow-through: contact that extended past the structured session itself.

What this means in practice: design the calendar before you design the content. Reduce week one to almost nothing. Give week two a memory. Hand week three a task. Use week four to check for contact outside the room, not just attendance inside it.

Week	Failure risk if skipped
1	High-friction entry deters return before content is judged
2	Group feels like a series of events, not a continuing place
3	Entire structure depends on facilitator presence
4	Programme optimises for attendance, not for connection outside sessions

What this does not solve

This sequence addresses retention among people who already showed up to week one. It does nothing for the people who never enrolled, and enrollment itself is skewed toward those with enough existing social capital, transport, and free time to attend a structured programme at all. Sequencing a first month well makes a good programme work better. It does not make an inaccessible programme accessible, and it is not a substitute for asking, separately, who is missing from the room in week one.