Co-production in Research for Charities

What happens when your research talks back?

Late nights spent polishing the plan, sharpening questions, ensuring ethics are watertight and the budget balanced. Yet a nagging doubt lingers. Your role is to amplify voices, yet you sense they might be limited, not deliberately, but by the process itself.

In charity research, achieving true representativeness is challenging. Populations are messy; skewed by age, location, language, tech access, work shifts, caring duties, confidence with forms, or simply who can make a Tuesday 10 a.m. slot.

Sticking to familiar channels often means hearing from the same voices. This risks some perspectives going unnoticed, not from disinterest, but because our approach quietly excludes them. Even 'representative' samples can blur sub-group nuances. For example, if an issue skews towards older groups, younger views may vanish in the aggregate. The overview seems complete but lacks the granularity needed for sharper frontline choices.

Governance keeps people safe and budgets keep us honest, but overly-tight briefs can inadvertently constrain discovery, missing vital questions that only those living it would think to ask. That’s where the extra happens: insight that lands closer to reality for strategy, participants who leave stronger than they arrived, and unplanned outputs that travel far beyond the original brief.

Below, we invite you to see co-production as a governance process, not a headache, and share practical ways to operationalise shared decision making power without chaos.

Quick Links — Jump to What Matters

• How can you make co-production research work?

• Practical lessons from the field

• What’s next: could ethical AI strengthen connection?

How can you make co-production in research work?

Co-production only works when it’s designed for real decisions, not just good intentions. Treat it as a governance process, with clear guardrails, defined stage-gates, and outputs that serve both communities and decision-makers. This helps research stay safe, focused, and accountable without closing the door on discovery. The below suggestions can help you put that into practice.

Open results needn't spell endless drift.

Guardrails: protect people and focus; name a safeguarding lead, use consent checkpoints, minimise data, and adopt a pause rule.

Stage-gates; decide with evidence, not drift: 3–5 criteria, booked dates, named decision-owners, explicit continue/pivot/stop.

Dual outputs; impact for leaders and community: a decision brief/dashboard and a participant-chosen community output, with clear permissions/credits.

Track success: What does good look like?

• Sub-group inclusion meets or exceeds population benchmark.

  Agenda share: ≥30% of agenda items from participants by phase 2.

• Gate discipline: decisions recorded on date (continue/pivot/stop + rationale).

• Capability gain: pre/post participant pulse ↑ in confidence/network/authorship.

• Community reach: community output hits a first-90-day use metric (plays/uses/adoptions).

Risks & how to mitigate

Public Project Examples: Practical Lessons from the Field

Shared for learning; inclusion doesn’t imply endorsement or affiliation.

These openly documented projects show different ways to make co-production practical. Each links straight to the original source. Click each for more info.

Common Ambition - Homeless health co-production

(Arch Health CIC, Justlife, University of Brighton; Health Foundation, 2021–24)

• Why it’s relevant: Equal partnership between people with lived experience, NHS providers and commissioners.

  
  

• Guardrails: Trauma-informed practice, clear group norms, paid involvement, external academic evaluation (University of Brighton), plus programme-level evaluation by the Health Foundation.

  
  

• Stage-gates: Local adoption of service prototypes (e.g. Route to Roof); training embedded in a medical school curriculum.

  
  

• Dual outputs: system maps, public workshops, practice resources.

Take-home: Co-chair steering groups, share ground rules and a pause rule, test a 90-day service with an adoption checkpoint.

Sources: BHCA case summary; BHCA final report; Arch Healthcare training page

Mind - Physical Activity Working Group

(national mental health charity, 2021–22)

• Why it’s relevant: co-produced campaign/resources led by people with lived experience of problematic exercise.

  
  

• Guardrails: wellbeing check-ins, clear boundaries, multiple feedback routes.

  
  

• Stage-gates: group-agreed aims; outputs iterated with lived-experience review.

  
  

• Dual outputs: practitioner guide plus first-person blogs/webinars that travelled widely.

Take-home: add a 30-minute wellbeing check-in to every workshop; commit to two outputs from day one.

Sources: Physical activity working group | Mind case studies; Co-production | Influence and Participation Toolkit

Common Ambition Bristol (HIV lived-experience partnership)

(Brigstowe, African Voices Forum, NHS partners; funded by the Health Foundation)

• Why it’s relevant: a lived-experience Project Delivery Group (community members & NHS clinicians) co-designed priorities and interventions to reduce HIV inequalities among African and Caribbean heritage communities in Bristol.

  
  

• Guardrails: multi-agency governance (charity, community network, NHS, public health, academia), clear group processes and shared decision-making roles.

  
  

• Stage-gates / decision impact: community-led outreach in barbershops and other trusted spaces; launch of dedicated community clinics; learning adopted into local strategy.

  
  

• Dual outputs: toolkit/report for replication; community resources (short film, radio, website) and a peer-reviewed research publication.

Take-home: co-chair your steering group with lived-experience partners; publish ground rules and a pause rule; run one 90-day pilot with an adoption gate; plan a community output alongside the decision brief.

Sources: NIHR ARC West (a); NIHR ARC West (b); Brigstowe - CAB clinics; Health Foundation

CAB videos (community-facing outputs); doi: 10.1136/sextrans-2025-056491

What’s next: can ethical AI help strengthen connection?

As these projects show, co-production thrives when we blend human insight with robust processes delivering not just sharper strategies, but ripple effects in communities. Looking ahead, the sector's next leap could lie in tools that amplify this without adding friction.

Ethical AI, handled with care, may offer a way to scale connections. Easing admin burdens, broadening access, and surfacing themes faster, all while keeping participants at the helm.

Imagine live captions unlocking workshops for non-native speakers, or auto-summaries sparking quicker validations, freeing researchers to focus on what matters most, the voices of lived experience. With the right guardrails, ethical AI could be an amplifier for impact.

Do less admin, free up space to truly listen

• Seamless entry: live captions, auto alt-text, plain rewrites, multi-language summaries; all human-vetted.

• Faster synthesis: auto-transcription with human-checked theming or condense sticky-note walls into themes for participant validation.

• Admin relief: scheduling, reminders, and action capture that becomes the shared participant agenda.

• Evidence visibility: turn literature into participant-friendly briefs.

Guardrails to add

• Ensure informed consent includes how AI tools are used.

• Data minimisation: redact identifiers by default.

• No model training on participant data; store outputs on approved/secure servers.

• Human-in-the-loop: participants (or an advocate) validate AI-summarised themes before any decision gate.

• Right to opt out: always offer a non-AI route.

• DPIA & ethics log: record tools, processors and retention; name a contact for concerns.

Success signals: Participant acceptance rate of AI-drafted themes; time saved on synthesis; accessibility upticks (e.g., caption/easy-read usage); leader rating of decision quality at the next gate.

Bringing it together

Incorporating genuine co-production into any research means you get insight that travels; decisions land closer to reality, participants leave stronger, and value shows up beyond a single PDF. The upside? It's doable under scrutiny. Guardrails, checkpoints, and twin outputs make participant-led research practical, trackable, and fundable right now. The public examples above show how, when done well, true co-production delivers benefits far beyond the original scope. Dipping into participant-led work? I hope this gives you practical guardrails to start safely.

Disclosure: The examples above are drawn from publicly available sources for illustrative and educational purposes only. Insights2Outputs has no current affiliations with the organisations mentioned. Any views are the author’s own.