Practice
Guidance index
Everything the Institute has published for practitioners. 220 documents, most recent first.
2026
The Recruitment Problem Social Media Data Just Made Harder to Ignore
A new nationally representative study on the closeness of online contacts is a reminder that recruitment channels which rely on people finding you tend to miss the people most worth reaching.
Where Attendance Actually Collapses, and What to Design Around It
Group-based connection programmes don't lose people gradually. They lose them at a specific point in the run. A look at how trial designs, social prescribing, and consumer platforms have each built around that point.
How to Choose Outcome Measures a Commissioner Will Actually Accept
A sequenced approach to picking loneliness and connection outcome measures that survive a funding review — not just the ones easiest to collect.
Matching the Intervention to the Driver, Not the Label
A referral marked 'lonely' is not enough information to act on. A step-by-step approach to working out what kind of disconnection you are dealing with before you choose an intervention.
What the Science Remote-Work Study Actually Gives You for a Cost Case
A new study puts a number on remote work's contribution to isolation. Here's what it does and doesn't let you claim to a finance director, and what evidence to use for the rest.
Training the Handoff Between Clinic and Community
Social prescribing depends on a moment neither health services nor community organisations are trained to manage well: the handoff. Here is what to train, and who, before you build the referral pathway.
Choosing Outcome Measures a Commissioner Will Accept
A framework for selecting loneliness and connection outcomes that hold up to commissioner scrutiny, built around which measures the evidence base actually supports.
The First Four Weeks: A Sequencing Guide for New Connection Programmes
Most community connection programmes lose people between week two and week four, not at launch. A week-by-week sequence for what to prioritise, and why befriending-only models keep underperforming structured activity.
Measuring the Handoff, Not Just the Outcome
How to evaluate a social prescribing referral pathway when the thing most likely to fail is the handoff between clinician and community link worker — not the activity at the end of it.
Writing the Business Case Commissioners Actually Fund
Commissioners do not fund loneliness reduction in the abstract. They fund service use, referral volume, and cost per outcome inside a budget cycle — and most social prescribing bids are written for the wrong audience.
Measuring Safety, Not Just Outcomes, in One-to-One Befriending
Social prescribing and befriending schemes measure loneliness scores and service-use reductions well. They measure safeguarding risk almost not at all — here is what the published evidence actually covers, and what it leaves for you to build.
How to Structure Roles Inside a Social Prescribing Group So It Doesn't Collapse by Week Four
A procedure for assigning roles inside a referral-based social group — host, anchor, and connector — so participation survives past the first few sessions.
Why Loneliness Predictors Should Decide Who You Recruit as a Volunteer
A new cross-national study on who is lonely and why has direct implications for befriending and social prescribing schemes: most volunteer pipelines recruit the wrong people for the job.
How to Recruit the People Who Won't Sign Up
A sequenced approach to reaching socially isolated people through programme design, when flyers, open calls, and self-referral only ever find the already-connected.
Where Social Prescribing Groups Actually Lose People
Retention in community groups and social prescribing programmes doesn't fail at the first session. It fails when the original reason for attending runs out and nothing has replaced it. A look at how three models handle that point differently.
When a Funder Asks for Evidence Your Programme Cannot Produce
A practice note for when a commissioner or funder wants your loneliness or social prescribing programme to demonstrate outcomes the intervention literature has not yet established. What to renegotiate, and how.
Isolated Is Not the Same as Lonely — Your Programme Design Needs to Know Which
A new study splits isolated older adults into those who become lonely and those who don't. That distinction should change how you screen, refer, and staff social prescribing caseloads.
Choosing Between a Digital and In-Person Group: What the Evidence Actually Supports
Before you build a group programme, decide whether it runs online or in a room. The evidence on which format reduces loneliness is thinner than most funders assume — here is how to make the call and measure it honestly.
Scaling Social Prescribing Without Losing What Made It Work
A practice note on the specific mechanism that breaks when social prescribing schemes grow, and what to protect deliberately rather than let erode.
Why the Workplace Social Programme Rarely Touches Loneliness
A post-mortem on the standard employer response to workplace loneliness data — optional social events layered onto unchanged work structures — and why it fails to move the numbers it was built to fix.
How to Decide When to Stop Running a Connection Programme
A sequenced procedure for deciding whether to close, redesign, or keep funding a social connection programme, built around the evidence you actually have rather than the evidence you wish you had.
2025
How Venue and Logistics Decisions Quietly Kill Attendance
A sequenced checklist for choosing venue, timing, and format for community programmes, covering the operational decisions that determine whether people show up twice.
What to Train Frontline Staff to Say About Loneliness — and What to Cut
A script audit for social prescribing link workers and clinicians: which phrases about loneliness and isolation are backed by evidence, which overpromise, and which quietly insult the person in front of you.
Evaluating a Workplace Connection Programme: What You Can Actually Measure
Most employers evaluate loneliness programmes with a single satisfaction survey. Here is a design that survives scrutiny without requiring a research team.