Should Your Group Meet Online, In Person, or Both

A commissioner asks why the new peer support group can’t just run on Zoom. Fewer no-shows from bad weather, no room hire, people can join from a hospital bed. All true. It is also the wrong first question. The right one is what the group is actually for — and that answer, not convenience, should decide format.

The trade you’re actually making

Digital delivery does not remove barriers to connection. It swaps physical barriers for a different set: screen fatigue, unequal comfort with video, the ease of muting yourself out of the conversation. Gallup’s 2024 workplace data gives a clean read on this trade-off in a comparable setting: fully remote employees report loneliness at 25%, hybrid workers at 21%, and fully on-site workers at 16%. That’s not proof video calls cause loneliness — remote workers differ from on-site ones in other ways too — but it is a consistent gradient across a large sample, and it should make you suspicious of any claim that format doesn’t matter.

The research on healthcare workers during remote shifts adds a useful distinction: workplace isolation and loneliness are separate constructs with different causes. A person can be well-supported and still isolated by the format itself. That’s the risk with an online group — everyone’s fine, and nobody is actually connecting.

Face-to-face socializing in the United States has been declining for decades, according to American Time Use Survey data reported by WBUR, and the decline predates the pandemic and continued after it. That’s the backdrop against which any group operates: you are not choosing between a healthy baseline and a digital substitute. You are choosing between two different responses to a pre-existing shortage of in-person contact, and only one of them adds any back.

What in-person adds that digital structurally cannot

Ray Oldenburg’s concept of the third place — informal space distinct from home and work — depends on incidental contact: the conversation that happens because you’re both standing at the counter, not because either of you scheduled it. A video call has no counter. Every interaction on it is scheduled, framed, and time-boxed. Eric Klinenberg’s argument about social infrastructure goes further: shared physical space doesn’t just host connection, it produces it, sometimes with life-or-death consequences, as in his account of differential survival during the 1995 Chicago heat wave. Nothing about a Zoom room replicates that.

There’s also a supply problem worth naming: the physical third places a digital group might otherwise have used are disappearing. A 2025 study in Health & Place tracked twelve categories of third place — coffee shops, libraries, museums, recreation centres, restaurants — and found closures across every category between 2019 and 2021, concentrated in socially vulnerable and rural areas. Reporting from the University of Colorado Boulder on the same research frames the loss of senior centres and churches as a driver of loneliness in its own right, not just a symptom of it. If you’re weighing digital against in-person, check first whether an accessible in-person venue exists in the area you’re serving. In some places it genuinely doesn’t, and that changes the decision.

Qualitative research on social prescribing offers a related finding: participants describe the benefit of group referral as extending beyond contact itself to restored purpose and participation, and structured, purposeful activity appears to work better than contact alone. That’s a point in favour of in-person formats generically — shared task, shared space — but it also tells you something about what a digital group needs to succeed if it’s your only option: it can’t just be a conversation. It needs a task.

When digital is the right call anyway

None of this means digital delivery is a lesser tier of service. It’s the correct choice, not a compromise, for:

1. Reach beyond a fixed catchment. A rare-condition peer group serving a rural region will never assemble a room. Digital is the only version of this group that exists.
2. Mobility or health barriers that in-person can’t solve. Post-surgical recovery groups, chronic pain groups, and groups for people who are housebound belong online, full stop.
3. A genuine on-ramp for people who would not attend in person yet. Some anxious or newly bereaved participants use a low-stakes digital session as a bridge to eventually showing up somewhere physical. Treat it as a bridge, not a destination, and say so to participants.

Where digital groups do work, the research on college students post-pandemic offers a clue why: what predicted wellbeing wasn’t the volume of social contact but its quality and regularity. A digital group that meets at a fixed time every week, with the same facilitator and the same small set of faces, can deliver that. A digital group that’s open-drop-in, different attendees every time, generally can’t — it has the structure of a webinar, not a group.

The “waiting room problem”

Name the failure mode so you can catch it: a digital group that opens with five minutes of silence while people’s cameras load, moves into a facilitator monologue, and closes the moment the hour is up, with no unstructured time before or after. Call this the waiting room problem — participants experience the session as something to attend, not somewhere to be. In person, the ten minutes before and after the formal agenda are often where the actual relationship-building happens; people arrive early to chat, linger to help stack chairs. Digital sessions default to cutting that time out because nobody wants to sit on a call with nothing scheduled. If you’re running digitally, you have to engineer that unstructured time deliberately — an open five minutes at the start with no agenda, a standing “stay on if you want to keep talking” at the end — or you’ve built a webinar with better intentions.

What this means in practice: don’t ask “online or in person” as a single global choice for the programme. Ask it per cohort, based on what’s stopping people from attending at all — distance, health, or comfort — and design the format to answer that specific barrier, not to save room hire.

Evidence status

Claim	Evidence status
Remote workers report more loneliness than on-site workers	Consistent large-sample survey finding; not causal
Physical third places produce connection that scheduled video calls don’t	Theoretical and observational, well-established in the literature, not tested experimentally against digital formats head-to-head
Purposeful, structured group activity outperforms contact alone	Reasonably consistent across qualitative social prescribing research
Regularity and quality of contact matter more than volume	Supported in recent post-pandemic student research; plausible generalisation, not yet a broad consensus finding
Digital groups without unstructured time underperform in-person equivalents	Practice inference, not directly tested; treat as a working hypothesis, not a settled fact

Worth flagging separately: commentary from George Mason University’s public health faculty on AI companionship tools raises the same underlying concern from a different angle — that a technically available substitute for human contact can quietly displace the harder, more effective option rather than supplementing it. The same caution applies to defaulting a group online because it’s easier to run, not because it’s what the participants actually need.

What this does not solve

Format decisions don’t fix a group that has no clear purpose, no consistent facilitator, or no plan for what happens after week six. They also don’t solve reach: whichever format you choose, you’re still designing for people who already agreed to show up somewhere — a physical room or a login link — and the group that most needs building, for people who won’t do either, isn’t answered by this note.