Making the Case to Commissioners: What to Measure and Why

A commissioner reading your bid does not want to know that loneliness is bad for health. They know that. What they want to know is whether the money they give you will move a number they can defend to their own board, their own auditors, or a scrutiny committee. That is a different question, and most programme evaluation plans answer the wrong one.

This guide is about closing that gap: what to measure, in what order, and how to talk about the limits of the evidence without talking yourself out of funding.

Start from what the commissioner has to justify, not what you can easily count

Commissioners in health and care systems are usually justifying spend against one of three things: reduced service demand, statutory duty, or a named strategic priority. Attendance figures satisfy none of these. A report that says “40 people attended eight sessions” tells a commissioner you ran something. It does not tell them whether that something was worth the alternative use of the money.

Work backwards from the commissioner’s own accountability line:

1. If they answer to a health board or ICB, they need something that plausibly connects to service use — GP contacts, A&E attendance, admissions. The systematic review of social prescribing and loneliness found that three of nine included studies reported reductions in GP, emergency, social worker, or inpatient use. That is a genuine finding, but it is drawn from a small and heterogeneous evidence base, so present it as “some studies show this,” not “social prescribing reduces GP visits.”
2. If they answer to a local authority scrutiny function, they are more likely to need well-being and independent-living measures, plus a plausible read on isolation among older residents — the National Academies’ concern that roughly a quarter of adults 65 and older are socially isolated gives you a population-level framing.
3. If they answer to a national strategy or public health target, cite the strategy itself. The UK’s 2018 loneliness strategy explicitly built loneliness measurement into national statistics; a commissioner working against that architecture wants to see your outcomes lined up with the same measures the strategy uses, not a bespoke scale you invented for this contract.

The three things worth measuring, and the one that isn’t

What you might measure	Evidence status	What to do with it
Attendance / retention	Descriptive only, no causal claim possible	Report it as delivery data, not outcome data. Never present it as evidence of impact.
Self-reported loneliness/isolation (validated scale)	Reasonable evidence social prescribing shifts self-esteem and confidence; loneliness-specific effects less consistently shown	Use a validated scale (UCLA-3 or similar), not a bespoke satisfaction survey, so a commissioner can compare across programmes
Service use (GP, A&E, admissions)	Weak: a few studies show reductions, evidence base is small and heterogeneous	Present as a hypothesis worth tracking over a funding period, not a promised outcome
Sense of purpose / meaningful participation	Increasingly seen as the mechanism, not just a side effect	Structured, purposeful activity appears to outperform unstructured contact — worth measuring directly rather than assuming contact alone delivers it

The item commissioners most want — a causal claim that your programme reduces NHS or council costs — is the item the evidence least supports. The 2025 systematic review protocol on social prescribing for older adults notes plainly that only one peer-reviewed randomised controlled trial exists in this space. One trial. Everything else commissioners have seen cited to them is observational, uncontrolled, or a qualitative synthesis. Say this out loud in your evaluation plan. A commissioner who has sat through enough funding cycles will trust you more for naming the gap than for papering over it with a confident cost-avoidance figure you cannot support.

What this means in practice: put the causal claim second, not first. Lead your evaluation section with a validated, comparable measure of loneliness or isolation and a clear account of delivery (who attended, how often, for how long). Put the service-use hypothesis further down, flagged as exploratory, with a plan to track it over multiple funding cycles rather than promise it in year one.

The “one clean trial” problem

Most of what gets cited to commissioners as evidence for social connection interventions is not from a randomised trial. That matters because commissioners increasingly sit opposite finance officers who have been trained to ask “compared to what?” — and an uncontrolled before-and-after loneliness score cannot answer that question. The HEAL-HOA trial testing volunteering against a control among lonely older adults in Hong Kong is notable mainly because it is rare: a genuine randomised comparison in a field dominated by small, uncontrolled programme evaluations. If your commissioner is sophisticated, they may ask why you are not running something closer to that design. The honest answer is usually cost and timeline, and that is an acceptable answer — but only if you say so rather than let a weak study stand in for strong evidence.

The failure mode: outcome-washing the logic model

The most common way evaluation plans go wrong here is not weak data — it’s a logic model that quietly upgrades correlation into causation between the funding application and the final report. A bid says the programme “will reduce loneliness and associated health costs.” Eighteen months later the final report says loneliness scores fell and GP contacts fell in the same period, and presents these as linked, when nothing in the design supports that link. Commissioners who catch this once do not fund that provider again. Name the mechanism you are testing, name the outcome you can actually measure, and keep a visible line between the two.

What this does not solve

This framework helps you choose defensible measures and talk to commissioners in terms they use internally. It does not fix the underlying evidence gap: outside a handful of trials, the field cannot yet tell a commissioner with confidence that a given social connection programme reduces cost of care. It also does not solve reach — outcome data from people who completed a twelve-week programme says nothing about the people who never referred themselves, and a commissioner reading only your headline figures will not see them either.