Digital or In-Person: How to Cost and Choose a Group Format

Most groups aimed at reducing isolation get their format decided before anyone asks whether it should be. A commissioner defaults to video calls because the last three programmes ran that way. A community organisation defaults to a room because that’s what the building is for. Neither decision gets tested against what the money is meant to achieve.

This is a funding decision, not a facilities decision. Format determines cost per participant, dropout rate, and — increasingly, the evidence suggests — whether the thing you’re funding actually works. Treat it as a step in the commissioning process, not an afterthought to the venue booking.

Step 1: Name the outcome before the format

Write down what you’re funding, in a sentence that could be measured. “Reduce loneliness among newly retired men in the borough” is a different funding problem from “increase attendance at a existing lunch club” or “give remote workers a reason to leave the house.” Each implies a different tolerance for digital delivery.

The distinction that matters most here comes from work on workplace isolation, which treats loneliness and isolation as related but separate constructs — one about subjective feeling, one about the structure of contact available to a person. A 2024 study of healthcare workers found perceived social support moderates how remote arrangements affect wellbeing: the same working pattern that isolates one person barely touches another who has support elsewhere. That has a direct funding implication. If your target population already has some social infrastructure and just needs more density of contact, a digital group can plausibly add value. If the population’s core problem is a lack of any social infrastructure at all, digital delivery adds a group to a life that already has none of the ordinary, unplanned contact that group is trying to substitute for.

Step 2: Look at what the workplace data actually shows

The workplace is the best-measured environment for this comparison, because employers have been running the natural experiment for years. Gallup’s 2024 State of the Global Workplace found that fully remote employees report loneliness at 25%, hybrid employees at 21%, and fully on-site employees at 16%. That is not a small gap, and it runs in one direction consistently across the gradient — more physical co-presence, less loneliness, in a dataset covering a global workforce.

This does not settle the question for community or clinical groups, which are a different kind of contact than colleague proximity. But it should raise the bar of evidence a funder demands before backing a digital-only model for a target population with limited existing social infrastructure, purely on cost grounds.

Step 3: Price the two models honestly

The comparison funders usually run is incomplete. It compares venue hire and travel against a video licence and staff time, and digital wins on paper every time. Run the full comparison instead.

Cost line	In-person	Digital
Venue / room hire	Yes	None
Facilitator travel time	Yes	None
Participant travel cost/barrier	Yes, and can exclude people	Minimal
Technology access and support	Minimal	Often underestimated — devices, data, digital literacy support
Facilitator skill required to sustain engagement	Moderate	Higher — video groups fail without deliberate structuring
Dropout and re-recruitment cost	Lower once attendance habit forms	Higher; digital attrition tends to be steeper
Incidental contact captured (queue, walk to the car, tea afterwards)	Free, and often the actual mechanism of benefit	Zero

That last row is the one budgets consistently miss. Ray Oldenburg’s account of the third place and Eric Klinenberg’s work on social infrastructure both make the same point from different angles: a lot of the benefit of a group happens in the unstructured minutes around it, not in the structured minutes of it. A video call has no equivalent of the walk to the car park. If your funding case rests on incidental contact as a mechanism — and for isolation-focused programmes it often should — digital delivery isn’t a cheaper version of the same intervention. It’s a different intervention with a missing ingredient, and the cost comparison needs to reflect that it may need to do more, not less, deliberate work to compensate.

Step 4: Check what the group actually needs to do

A 2022 meta-synthesis of participant experience in social prescribing found that structured, purposeful group activity delivered more benefit than unstructured contact alone. People described the value as extending beyond social contact to a sense of restored participation and purpose. This is useful for the format question because it suggests the deciding factor is not physical presence per se, but whether the activity gives people something to do together, not just something to look at each other while doing.

That reframes the digital-versus-in-person choice usefully. A digital cooking demonstration where everyone follows along in their own kitchen has a task. A digital discussion group with six video tiles and no shared activity does not, and this is where digital groups tend to fail — call it the gallery-view problem, six faces on a screen with no shared task holding the session together, and attention drains within twenty minutes.

What this means in practice: before committing to digital delivery, check whether the format can carry a shared task, not just shared attendance. If the answer is no, either build one in — a shared document, a physical object everyone has been posted in advance, a structured round — or fund the in-person version instead, because the cost saving will be eaten by dropout.

Step 5: Decide, then pilot small before scaling

Once you’ve costed both models against the actual outcome and checked whether the activity survives a screen, make the call and pilot it at a scale you can afford to be wrong at — six to eight weeks, one cohort, before committing a full grant cycle. The American Heart Association’s 2022 scientific statement on social isolation makes a point that applies directly to funding decisions: the intervention evidence base is thin. The statement identifies the absence of trial-quality evidence on what actually reduces isolation as the central research gap in the field. No one, digital or in-person, is funding from a position of settled proof. That is a reason to measure your own pilot properly, not a reason to skip the comparison.

Evidence status

Claim	Status
On-site work is associated with lower loneliness than remote work	Reasonably well supported, large workplace dataset
Structured purposeful activity outperforms unstructured contact	Supported by qualitative synthesis, not randomised trials
Incidental contact around a group adds to its benefit	Theoretically grounded, not directly measured against digital equivalents
Digital groups can be as effective as in-person for isolation reduction	Genuinely unsettled — no cited intervention trial resolves this either way

What this does not solve

None of this tells you how to reach the person who won’t join a group in any format — the housebound, the recently bereaved, the person too anxious to attend the first session. Format choice is a second-order decision that only matters once someone has already decided to show up. The bigger funding problem, reaching people who wouldn’t otherwise seek a group out, sits upstream of everything in this piece.