Our Approach
Most research asks what happened. We also ask who was allowed to decide it.
Four phases, and who holds the decision in each.
Every research process has phases. What differs is who is allowed to change the answer once the work is underway.
- Phase 01, Discover and Scope. What happens: We listen first. Context, aims, stakeholders, and an approach shaped to the actual constraints rather than a template. Decision held by: The commissioner and the community, jointly. Scope is agreed with people from the community before it is signed off. Control point after this phase: Community researchers recruited, trained and paid.
- Phase 02, Co-design and Build. What happens: Methods, instruments and ethics. Community researchers are recruited and trained here, not later. Decision held by: Community researchers. They shape the questions and they approve the coding frame before any analysis begins.
- Phase 03, Evidence in Practice. What happens: Fieldwork in trusted spaces, with quality assurance and safeguarding running alongside it. Decision held by: Participants. Consent is ongoing rather than a signature at the start, and anyone can withdraw what they said. Control point after this phase: Findings returned before they are reported.
- Phase 04, Sensemaking and Action. What happens: Analysis, triangulation, and the sessions where partners have to agree what the evidence means. Decision held by: The community first, then the board. Findings are returned to the people who gave them before they reach a commissioner.
The frameworks are the product.
Anyone can run a focus group. What makes findings usable is the structure you assess them against, and we build ours. These are ours, in the form a commissioner would need to assess them.
Trust metrics
Trust is the thing every health programme claims to be building and almost nobody reports on. We built a way to describe it that survives contact with a programme board: comparable over time, honest about direction of travel, and specific enough to act on.
What it answers
- Has trust moved, in which direction, and for whom.
- What was done that moved it, and what was done that cost it.
- What a board can change inside the current funding cycle.
Where it has been used
Embedded in a live governance evaluation of a multi-partner health programme.
Power-sharing criteria
Four tests applied to a decision-making structure rather than to a set of intentions. Distribution, accountability, visibility and adaptability. Applied across every tier of a programme, from the board down to delivery partners, they show where authority actually sits rather than where the terms of reference say it sits.
What it answers
- Who set the agenda, who could escalate, and who was told afterwards.
- Whether a stated commitment to sharing power is visible in any decision record.
- Whether the structure is capable of changing, or only of consulting.
Where it has been used
A live governance evaluation across a multi-tier programme structure.
Community researcher model
Recruitment, training, supervision and quality assurance for researchers drawn from the communities a study is about. Paid work, not participation. This is the part most organisations describe as co-production and resource as a focus group.
What it answers
- Who is asking the question, and why that changes the answer.
- How quality is assured without defaulting to academic gatekeeping.
- What the researchers take with them when the contract ends.
Quality framework
- Validity.
- Methods appropriate to research questions
- Reliability.
- Consistent, replicable approaches
- Reflexivity.
- Awareness of researcher position and bias
- Transparency.
- Clear documentation and audit trails
- Feedback loops.
- Validation with participants and stakeholders
Where it has been used
More than five years training people across London, and running fieldwork on a live maternal health programme in south east London.
Analysis at scale
We use computational methods to handle volume. Community researchers decide what the data means. Machine assistance does a first pass on large qualitative sets, agreement between coders is reported rather than assumed, and the coding frame is reviewed by the community before analysis proceeds.
What it answers
- How a large evidence base is made navigable without flattening what is in it.
- Where a machine was used, on what, and what a human checked afterwards.
- Who controls the categories the evidence gets sorted into.
Where it has been used
Applied to large qualitative datasets and to service and outcomes data alongside them.
Values that guide every project
These principles are not aspirations. They shape our methods, our team composition and our relationships.
Culturally grounded
We design approaches that honour cultural context, recognising that effective research must be responsive to the communities it serves.
Safe by design
Every interaction is designed with awareness of harm. We prioritise participant wellbeing and create safe spaces for sharing.
Anti-racist practice
We actively work to identify and address racism in research processes, from who asks questions to who benefits from answers.
Transparent and reflexive
We are honest about our position, our limitations and what we have learned. Openness is how trust gets built.
Ethics and governance
Rigorous governance underpins everything we do.
Ethics and safeguarding
Formal ethics processes, safeguarding protocols, informed consent, and clear complaints procedures.
Data governance
GDPR-compliant data handling, secure storage, clear data sharing agreements, and transparent retention policies.
Questions commissioners ask us
The things worth knowing before a brief is written. If something is missing, ask us directly.
What does Urban Insight Collective do?
We are a community research and evaluation consultancy. We design and deliver research, evaluation, co-design and evidence synthesis for the NHS, local authorities, foundations and voluntary sector organisations. The work is community-led and held to academic standards of rigour.
Where are you based and where do you work?
We were built in Brixton and we work across London. We work from the boroughs the evidence is about rather than commenting from a distance.
What methods do you use?
Mixed methods. Qualitative fieldwork including interviews, focus groups and ethnographic observation, quantitative survey design and analysis, participatory co-design, and computational analysis of large qualitative sets. Findings are triangulated and carry an audit trail from fieldwork to conclusion.
Do you use AI in your analysis?
Yes, and we say where. Computational methods do a first pass on large qualitative sets to handle volume. Agreement between coders is reported rather than assumed, and the coding frame is reviewed by the community before analysis proceeds. Community researchers decide what the data means.
How do you involve community researchers?
Community researchers are recruited, trained, supervised and paid as part of the delivery team. They are on the payroll rather than in the appendix, and they take part in analysis and sensemaking, not only data collection. This is set out in full in the community researcher model.
How do you handle ethics, safeguarding and data protection?
We run formal ethics processes, safeguarding protocols, informed consent and a clear complaints route. Data handling is GDPR compliant, with secure storage, data sharing agreements agreed at the start of a project, and transparent retention policies.
How does a project usually run?
In four phases: discover and scope, co-design and build, evidence in practice, then sensemaking and action. We state who holds the decision in each phase so accountability is explicit from the start.
What do you deliver at the end of a project?
Evaluation frameworks and logic models, insight and evaluation reports, evidence catalogues and maps, learning briefs, stakeholder packs, interactive explorers, toolkits, and training for community researchers. The exact package is agreed during scoping.
Can you work as part of a consortium or as a subcontractor?
We work as lead supplier, and we deliver alongside partner organisations where a programme needs more than one specialism. Tell us the shape of the work and we will be straight about whether we should lead it, partner on it, or step back.
How do we start a conversation?
Book a thirty minute call or email info@urban-ic.com. It is useful to talk before the brief is written, and we will say so if we are not the right people for the work.