Writing a Connection Programme Bid Commissioners Will Actually Fund

A commissioner reading a bid for a social connection programme is not moved by “loneliness is a public health crisis.” They have read that sentence in every bid this quarter. What they are actually scanning for is whether your programme reduces something they are accountable for — GP attendances, A&E presentations, sickness absence, care package costs, safeguarding referrals — and whether you can tell them, credibly, when they will see it.

Most bids fail this not because the programme is weak but because the writing answers a question nobody asked. It leads with epidemiology and ends with an aspiration. Commissioners need the reverse: the outcome first, the mechanism second, the epidemiology as one line of context if at all.

This is a framework for making that switch. It will not fix a weak programme. It will stop a decent programme losing to a worse one that happened to be written in the commissioner’s language.

What the commissioner is actually deciding

Strip away the specific outcomes framework and a commissioner is answering four questions, in this order:

1. Does this reduce demand on something I’m accountable for, or does it just make people feel better? Both matter, but they are funded from different budget lines and judged against different targets.
2. What is the unit cost, and what is the comparator? Not “loneliness costs the economy X” — that number, wherever it comes from, doesn’t tell them what your specific programme costs per person reached, per outcome achieved.
3. How does a person get in and out? Referral pathway, eligibility, dosage, exit criteria. A programme with no defined exit reads as an open-ended cost.
4. What will you tell me at month six, and will it be believable? Not whether you’ll report — everyone promises that — but whether the metric you’re proposing is one they can defend to their own board.

Your bid should answer these in this order. Most bids answer them in reverse, or not at all.

Translating the evidence base into commissioner terms

The academic literature on loneliness and isolation is now large enough that citing it well is table stakes, but most bids cite it badly — as scale-of-problem framing rather than as decision-relevant evidence. Here is roughly what each major finding is actually good for in a bid.

Finding	What it’s good for in a bid	What it can’t do
Holt-Lunstad’s 2010 meta-analysis: weak social relationships carry a mortality risk comparable to established clinical risk factors	Establishing that connection belongs in a health outcomes framework, not a wellbeing add-on	Does not tell a commissioner your programme changes mortality — no programme evaluation has shown that
Holt-Lunstad’s 2015 review: isolation, loneliness, and living alone each independently predict earlier death, more strongly under age 65	Justifying working-age as well as older-adult cohorts	Weak on dose — doesn’t specify how much contact reverses the risk
The American Heart Association’s 2022 statement linking isolation to a roughly 30% higher risk of heart attack, stroke, or death from either	Strong for cardiovascular-linked commissioning conversations, especially with integrated care boards	The statement itself says the intervention evidence is the gap — do not imply your programme closes it
The 2023 Surgeon General advisory’s finding that around half of US adults report loneliness	One line of context, not a case for scale	Says nothing about who is reached by any given programme design
The National Academies’ 2020 finding that roughly a quarter of adults 65+ are socially isolated	Useful for eligibility criteria and target-population sizing	Isolation and loneliness are distinct; don’t conflate them in your outcome claims
Systematic reviews of social prescribing (2021, 2021) reporting self-esteem and confidence gains, and reduced GP/A&E use in some included studies	The closest thing to service-use evidence you have — use it, cautiously	Both reviews flag small trial numbers and heterogeneous designs; don’t state the service-use finding as settled
Qualitative synthesis (2022) finding structured, purposeful activity outperforms unstructured contact	Justifies programme design choices — activity with a purpose, not “drop-in social space”	Qualitative; can’t support a quantified claim
The 2023 BMC review noting inconsistent measurement across the loneliness literature	Use defensively — it’s why you’re proposing one specific validated tool rather than a bespoke survey	Doesn’t help you if your own outcome measure is inconsistent too

What this means in practice: lead your evidence section with the AHA cardiovascular statement or the social prescribing service-use findings if your funder is a health commissioner. Lead with the National Academies isolation prevalence if you’re bidding into an adult social care budget. Do not lead with the Surgeon General advisory’s headline prevalence figure — every commissioner has already seen it, and it establishes scale, not case.

The failure mode: the epidemic pitch

Call it the epidemic pitch. It’s the bid that opens with the size of the problem — a third of adults, half of Americans, a public health crisis on the scale of smoking — and then, three paragraphs later, proposes a weekly walking group for twelve people. The gap between the scale of the framing and the scale of the intervention is what kills it. Commissioners read hundreds of these. They have learned to discount the opening paragraph and skip to the budget line, and by the time they get there they’ve already decided the bid is generic.

The fix isn’t a smaller epidemiology claim. It’s matching your framing to your actual reach. If you’re running a twelve-person walking group, say what it does for those twelve people, what it costs, and what population segment it’s targeted at — then let the commissioner draw the connection to the wider problem themselves. They already know the epidemiology. What they don’t know is why your twelve-person group is the right unit of intervention for the referral cohort they’re worried about.

A four-part structure that answers the commissioner’s actual questions

1. Outcome and comparator, in one paragraph. State the primary outcome in the commissioner’s own metric — reduced GP contacts, reduced repeat safeguarding referrals, reduced sickness absence days — and name what you’re comparing it against, even if the comparison is “no structured intervention” rather than a randomized trial. Cigna’s 2020 finding that lonely workers report missing work roughly five times as often due to stress is useful here for a workplace commissioner, but state it as context for the outcome you’re targeting, not as the outcome itself.

2. Mechanism, briefly. One paragraph on why this specific design produces that outcome — referencing the social prescribing evidence on structured, purposeful activity outperforming unstructured contact, if that’s your design logic. Don’t restate the full literature; cite the one finding that justifies your specific design choice.

3. Pathway and dosage. How someone is referred in, how long they stay, and what “done” looks like. Commissioners fund things with an end state. An open-ended drop-in with no defined exit reads as recurring cost with no ceiling, regardless of how good the activity is.

4. What you’ll report and when, in their language. Not “we will evaluate wellbeing outcomes” — name the instrument, the timepoints, and be honest about what it can and can’t show at that scale. If you’re running a twelve-person pilot, say plainly that it will generate a service-use signal, not a causal claim, and that the causal claim requires scale you’re not proposing yet. Commissioners fund honest pilots more readily than they fund overclaimed ones, because the overclaim is the thing that gets a programme cut at year two when the promised effect doesn’t show up in the data.

What this does not solve

None of this makes a mediocre programme fundable, and it does not substitute for a genuine theory of change. It also does not solve the underlying evidence gap: even the strongest cited sources here — the AHA statement most explicitly — state directly that intervention evidence lags behind the risk evidence. You are asking commissioners to fund a plausible mechanism supported by association data, not a proven cure, and a bid that pretends otherwise will eventually be caught out by its own evaluation.

And translating your pitch into commissioner language does nothing for the people the pathway never reaches. Referral-based programmes, by construction, reach people already in contact with a GP, a link worker, or an employer’s occupational health service. Whatever you write, and however well it’s funded, this framework is about winning the commissioning conversation for programmes serving people already inside a system — not about the harder problem of reaching the people who aren’t.