The Recruitment Problem Social Media Data Just Made Harder to Ignore

A nationally representative study published on PubMed on April 10 asked a simple question: does it matter whether the people someone talks to on social media are close contacts or distant ones? The study set out to test whether closeness moderates the relationship between social media contact and loneliness — not just whether people use social media, but who they’re using it to reach.

You don’t need the effect size to see the operational point. People whose online ties skew weak — acquaintances, people they follow rather than know, contacts who never become anything closer — are not the same population as people with strong online ties. And if your programme’s recruitment channel is social media, you are reaching whichever of those two populations shares, comments, clicks, and replies. That is very unlikely to be the more isolated one.

This is not a new problem. It is the same problem every self-selecting recruitment channel has. The new study just gives it sharper teeth, because it shows the mechanism operating inside the exact channel most programmes now default to for outreach.

The RSVP filter

Call it the RSVP filter: any recruitment method that requires a person to notice a flyer, open an email, follow a link, or fill in a form has already screened for people who have enough social bandwidth, trust, and forward motion to do those things. That is precisely what people experiencing serious isolation tend to lack. The 2018 AARP survey of adults 45 and older found that 61% of people who had never spoken to a neighbour were lonely, against 33% of those who had — loneliness correlates with the very small acts of initiation that most recruitment funnels assume will happen on their own.

Social prescribing research runs into this constantly. Qualitative work on how people experience social prescribing has found that benefit tends to come from structured, purposeful activity rather than contact alone — but that finding presupposes the person got through the door in the first place. A systematic review of social prescribing and loneliness found positive impacts across all nine included studies, but the populations in those studies were, almost by definition, people who had already engaged with a GP or link worker enough to be referred. That is a lower bar than social media self-selection, but it is still a bar.

What proactive recruitment actually looks like

The most useful recent evidence on this doesn’t come from a loneliness campaign at all — it comes from a trial design. The 2026 HEAL-HOA behavioural activation trial worked specifically with older adults living in poverty, living alone, and digitally excluded — a population that would never surface through a web form or a social post. Recruitment and delivery were both engineered around that fact: contact was by telephone, sessions were delivered by trained laypeople who were themselves older adults who had experienced loneliness, and the model required no digital literacy, no travel, and no self-initiated sign-up beyond answering a phone call. It reduced loneliness at 12 months against a befriending control.

Compare that with the earlier HEAL-HOA trial on volunteering and prosocial engagement, or the 2025 randomised trial of befriending in aged care, which reduced loneliness scores at 8 and 16 weeks. Both interventions work. But both, in their standard form, depend on someone being willing and able to opt in to a volunteering scheme or a befriending referral. The 2026 trial’s contribution to recruitment practice is not the behavioural activation content — it’s the delivery model that reached people who would not have reached it.

What this means in practice

If your only funnel into a programme requires the person to find you, you are recruiting the least isolated version of your target population. Build at least one channel that goes to people through an existing relationship they didn’t have to initiate — a GP referral, a housing officer, a school contact, a phone call from someone who already has their number on a caseload. Digital and self-referral channels can run alongside this, but they cannot be the whole strategy if reach to the most isolated is the goal.

Physical infrastructure matters here too, not just outreach mechanics. Research on third-place availability across the United States found closures in every category of informal gathering spot between 2019 and 2021, concentrated in socially vulnerable and rural census tracts — the same places where a telephone call from a trusted local contact will do more work than a poster in a library that no longer exists. Analysis of the same data has framed the loss of senior centres and churches as a driver of isolation, not a symptom of it, which matters for where you put a proactive recruitment effort: not just who you call, but whether there is anywhere to invite them to.

Evidence status

Claim	Status
Weaker online ties correlate with higher loneliness	Supported — new nationally representative study, closeness as moderator
Social prescribing helps once someone is referred	Supported, weak trial base
Proactive, non-digital outreach (telephone, trusted intermediary) reaches more isolated people than self-referral	Supported by trial design, not yet head-to-head tested against digital recruitment
Digital recruitment channels systematically undercount the most isolated	Plausible mechanism, not directly trial-tested

What this does not solve

None of this fixes the underlying supply problem: even a well-designed proactive recruitment channel still requires someone with a caseload, a phone line, and time to make the calls, and most funded programmes do not have spare capacity for that kind of legwork. A telephone call from a trusted intermediary reaches further than a social media post, but it is also far more expensive per contact. The honest constraint is not knowing that proactive outreach works better — it’s whether anyone is funded to do it at the scale the isolation actually requires.