Rethinking Co-Design for the Green Transition: Balancing Stakeholder Input and Designer Agency

Abstract

Co-design plays a pivotal role in architectural design and urban planning for the green transition, facilitating collaboration among designers and stakeholders to create contextually appropriate solutions. This study examines the balance between stakeholder input and designer agency within co-design practices aimed at addressing the complex challenges posed by the green transition. Looking at how designers’ mindsets and methods are influenced by co-design, this study is carried out by analysing two contrasting case studies from the Future Island-Island project: Field Operations, an immersive residential on Rathlin Island, and DesignLink, a structured design sprint with organisational partners. Employing the terminologies of autogenic (designer-led) and allogenic design (stakeholder-led), the research critically explores how these modalities influence design outcomes and designers themselves. Field Operations exemplifies a more allogenic approach characterised by collaborative brief development through local immersion, while DesignLink primarily illustrates an autogenic process where predefined objectives guided creative synthesis. The study reveals that effective co-design requires oscillation between these approaches, underscoring the necessity for designers to harness both community insights while ensuring their own creative agency. The findings in this study advocate for a refined co-design framework that optimally integrates stakeholder contributions without compromising the integrity and coherence of the design process, emphasising the importance of contextual sensitivity, innovation, and timely decision-making in addressing complex societal challenges such as the green transition.

1. Introduction

Co-design is commonly understood as a collaborative design process in which stakeholders actively contribute to the development of solutions or design responses [1]. Within the context of architectural design and urban planning, co-design has been lauded for its potential to respond to messy, complex challenges such as climate adaptation and community resilience [2,3,4]. Given the increasing urgency of the green transition requirements, co-design processes are essential for developing adaptive strategies that engage communities in resilience planning.

Despite the growing interest in co-design as a means to address sustainability and climate challenges, there remains a notable gap in methodological clarity regarding how designer agency is calibrated within participatory processes under the pressures of a green transition. This paper addresses that gap by comparing two contrasting formats within a single programme: an allogenic, place-based residency (Field Operations) and an autogenic, organisation-partnered design sprint (DesignLink). Through this comparison, we aim to propose a calibrated model for balancing stakeholder engagement with designer agency in contexts of environmental transition. The study contributes (i) an explicit definition and operationalisation of designer agency, (ii) a comparative case study design with transparent methodological reporting, and (iii) an evaluative framework with indicators spanning approach, process, and outcomes.

In this study, we define designer agency as the situated capacity of designers to shape the trajectory of a co-design process through three interrelated dimensions: (a) problem-framing—the ability to set, refine, or reconfigure the definition of the design challenge; (b) creative synthesis and decision-making autonomy—the authority to generate and select design directions; and (c) process stewardship—the responsibility to pace, structure, and guide collaborative work toward convergence. This definition acknowledges that agency is neither absolute nor fixed; rather, it is distributed, negotiated, and contingent on the dynamics of stakeholder engagement. By operationalising designer agency in these terms, we provide a clearer analytical lens for examining how different co-design formats enable or constrain the designer’s role within the broader participatory ecosystem.

Co-design has been widely adopted across international contexts, from the built environment to health and digital innovation, yet its methodological foundations remain uneven. In architecture, studies have highlighted both the potential and the limitations of co-design in creating inclusive environments, such as work on universal design and the participation of people with disabilities in public buildings [5], and the challenges that emerge in practice when multiple actors and agendas intersect [6]. Other contributions emphasise the transformative effects of co-design for participants, showing how processes foster ownership, empathy, and long-term commitment [7]. At the same time, evaluations of co-design are often ad hoc, prompting calls for more systematic assessment frameworks [8]. Beyond architecture, research in areas such as child–computer interaction and digital health demonstrates the adaptability of co-design to new domains, while also underscoring the need for clearer methods and indicators [9,10]. This broader landscape situates the present study within a growing international debate about how co-design can be both conceptually rigorous and practically effective in times of transition.

Evidence suggests that sustained engagement with local communities can yield outcomes that are more contextually appropriate, socially accepted, and environmentally sensitive [11]. However, while participatory approaches can enrich planning, they also present challenges. Gathering and reconciling input from diverse participants is often resource-intensive, demanding considerable time and funding [12]. Designers are uniquely positioned to synthesise these diverse inputs into coherent visions and prototypes through professional practice and training. A purely allogenic process, where decisions are entirely externally driven, may inhibit the creative autonomy necessary for innovation and natural creativity. Conversely, a fully autogenic process risks disregarding lived experience and other considerations such as contextual sensitivity or heritage.

This argument is explored qualitatively through two contrasting case studies from the Future Island-Island project: Field Operations, a five-day place-based residential on Rathlin Island; and DesignLink, a structured, three-day design sprint with organisational partners. The comparative analysis of these events illustrates the necessity of oscillating between allogenic and autogenic approaches across different project phases, creating a balance of approaches. The paper first outlines the research design and methodological approach, then presents findings from the two contrasting case studies, followed by a cross-case discussion of their implications. We conclude by proposing an evaluative framework with indicators for assessing co-design effectiveness and by reflecting on the relevance of these insights for future applications in the context of the green transition.

2. Methods

2.1. Research Design and Rationale

This study adopts a comparative multiple-case study design, a strategy well suited for architectural and design research where the aim is to investigate complex, context-dependent processes [13]. Case study research allows for in-depth examination of how co-design unfolds in real-world settings and enables cross-case synthesis to identify patterns, contrasts, and transferable insights.

The research was situated within the Future Island-Island programme, which offered two distinct formats of co-design engagement. We selected these cases through a maximum-variation logic: Field Operations represented an allogenic format, privileging external stakeholder voices through an intensive, place-based residency, while DesignLink represented an autogenic format, foregrounding designer agency within a structured design sprint with organisational partners. This contrast in agency distribution was chosen to test how different participatory structures affect outcomes in the context of the green transition.

Each case was treated as a bounded unit of analysis, with embedded units including design teams, stakeholders, and the artefacts they produced. This design enabled both within-case examination of processes and outcomes, and cross-case comparison to assess how oscillations between allogenic and autogenic approaches influence co-design effectiveness.

The comparative case study approach was supplemented with a qualitative analysis pipeline (described in Section 2.5) to ensure methodological transparency and rigor. Multiple sources of evidence were collected, including focus groups, facilitator notes, surveys, and design artefacts, and were systematically triangulated to strengthen validity. Following evaluation frameworks in co-design research [8], we structured our analysis around three dimensions: approach, process, and outcome, which guided both data collection and assessment.

2.2. Clarifying Future Island-Island

Future Island-Island is an AHRC-funded interdisciplinary research initiative focused on developing strategies to mitigate and adapt to climate change impacts within Northern Ireland, fostering sustainable practices through community engagement and design thinking. Operating across multiple work packages, the project tests models of regenerative innovation rooted in place-based knowledge, community engagement, and design thinking. Its relevance extends beyond regional development, contributing to broader discourses on sustainability transitions and participatory innovation models in the UK and internationally [14,15].

2.3. Design Methodologies Referenced

This research adopts the terminology of autogenic and allogenic design from ecological theory (see Appendix A), repurposed here to describe differing modalities within co-design practice. These terms offer a complementary lens to existing frameworks such as Sanders and Stappers’ [16] co-design landscape and Arnstein’s [17] ladder of participation, helping to articulate not only who participates, but also how agency is distributed across the design process.

Allogenic design refers to processes shaped predominantly by external agents, such as users, communities, or environmental contexts. Designers in this mode act as facilitators or convenors, with decision making emerging collaboratively and responsively. This approach aligns with the upper rungs of Arnstein’s [17] ladder, such as ‘partnership’ and ‘delegated power’, as well as with participatory design traditions that centre on local knowledge, relational ethics, and trust-building [18,19]. Within Sanders and Stappers’ [16] framework, this mode maps closely to the co-design and co-creation quadrants, in which end-users become active design partners rather than passive subjects.

Autogenic design, by contrast, is driven primarily by the internal logic of the design team or organising body. It entails a structured process, such as a design sprint or studio charrette, where objectives, roles, and timelines are predefined and purpose-made. While still informed by stakeholder insights, this modality prioritises internal creative synthesis, rapid iteration, and outcome-oriented development. It aligns more with the lower to mid-rungs of Arnstein’s ladder (e.g., “consultation” or “informing”) and sits closer to the design-led or expert-driven end of Sanders and Stappers’ spectrum. However, it is particularly effective when addressing urgent challenges requiring decisive action or innovation within tight timeframes [20].

Field Operations primarily exemplified the allogenic mode: co-created briefs emerged through immersion, listening, and self-organised collaboration with local residents. DesignLink, by contrast, operated under a predominantly autogenic framework: pre-scoped briefs and structured timeframes drove creative outcomes, albeit with moments of stakeholder feedback. Together, the two cases function as methodological counterweights, enabling reflection on when and how to modulate between participatory openness and designer agency.

Overlaying the autogenic/allogenic distinction onto existing models, such as Sanders and Stappers’ co-design landscape [16], offers a way to analyse temporal shifts and authorship dynamics—that is, not solely whether users are involved, but when, to what degree, and with what consequences. This overlay (Figure 1) has been visualised as a gradient cutting diagonally across their model, illustrating how projects might dynamically shift between allogenic and autogenic tendencies depending on the phase, urgency, or nature of the design challenge.

However, this approach does not provide a precise basis for comparison because the categories do not align with the focus of this investigation. To address this, modifying Sanders and Stappers’ diagram by substituting ‘stakeholder as originator’ for ‘user as subject/partner’ and ‘design as originator’ for ‘led by design/led by research’, with varying high and low degrees of accuracy, could offer a more accurate framework (Figure 2).

By contributing this ecological metaphor, we aim not to supplant existing co-design frameworks but to expand them—foregrounding how design agency itself can evolve in context, particularly in collaborative processes concerned with sustainability, power asymmetries, and systems-level change.

2.4. Data Sources and Collection

Data were collected from both case studies through multiple sources to enable triangulation and ensure a comprehensive account of each process, as can be seen in

Table 1. Summary of the primary data sources and their coverage across the two case studies.

Data Source	Field Operations	DesignLink
Participant focus groups	35 participants	23 participants
Surveys	N/A	Pre- and post-sprint surveys
Facilitator notes	Daily reflective notes	Daily reflective notes
Design artefacts	Visuals, prototypes, sketches	Concept proposals, briefs, visuals
Documentation	Photos, activity logs	Workshop recordings, transcripts

. For Field Operations, materials included daily facilitator notes, participant focus groups, photographic documentation of activities, and the design artefacts produced by teams (See Appendix B). For DesignLink, data comprised pre- and post-sprint surveys, workshop recordings, facilitator notes, and the outputs generated in response to organisational briefs (See Appendix C).

To maintain comparability across the cases, all material was anonymised and stored in a common repository. Consent was obtained from participants at the outset of each event, with particular care given to protecting the confidentiality of community representatives and organisational partners.

2.5. Analysis Pipeline

The analysis followed a multi-stage process designed to ensure transparency, comparability across cases, and robustness of interpretation.

1.

Preparation

All transcripts, notes, and survey responses were collated and anonymised. Materials were imported into a qualitative analysis environment to facilitate systematic coding through keywords.

2.

Theme development

Keywords were clustered into higher-order themes within each case, such as team organisation, stakeholder involvement, and outcome orientation. Case memos were drafted to capture emergent interpretations.

3.

Cross-case synthesis

Themes were compared across the two cases using a case–theme matrix to highlight contrasts between allogenic and autogenic structures. This process allowed us to trace how designer agency and stakeholder engagement shifted in different formats.

4.

Triangulation

Thematic findings were cross-checked against multiple data sources, including facilitator notes, survey responses, and design artefacts, to strengthen validity. Where possible, outcomes were further verified by follow-up conversations with organisers.

5.

Evaluation lens

To align with existing scholarship on assessing co-design [8], findings were mapped against three dimensions: approach, process, and outcomes. This ensured that both procedural and result-oriented aspects were considered in the analysis.

2.6. Evaluation Criteria and Indicators

To provide a transparent basis for comparison, we developed a set of evaluation criteria aligned with established approaches to co-design assessment [8]. Indicators were grouped into three dimensions: approach, process, and outcomes.

• Approach

• Clarity of project brief and objectives
• Visibility of stakeholder roles and responsibilities
• Alignment between chosen format and stated aims

• Process

• Timing and rhythm of stakeholder engagement (e.g., frequency and duration of interactions)
• Evidence of designer agency in framing, synthesis, and facilitation
• Degree of collaboration within and across teams

• Outcomes

• Tangible artefacts produced (e.g., concepts, prototypes, briefs)
• Relational outcomes such as trust, commitment, or willingness to continue collaboration
• Immediate follow-up actions (e.g., requests for implementation, funding applications, pilot initiatives)
• Indications of longer-term impact or sustainability of results

Both qualitative and quantitative data were used to populate these indicators. For example, survey responses were analysed to provide descriptive statistics on participant satisfaction, while focus group transcripts and facilitator notes were coded for evidence of stakeholder voice, perceived agency, and process effectiveness. Design artefacts and follow-up activities were used to assess implementation potential.

This framework ensured that evaluation went beyond descriptive narrative and offered a structured basis for interpreting the effectiveness of the two contrasting co-design formats.

2.7. Field Operations: A Place-Based, Emergent Co-Design Residential

Field Operations was a five-day design residency on Rathlin Island in October 2024 (timetable can be seen in Figure 3). As part of the ‘Future Island-Island’ initiative, its purpose was to collaboratively envision regenerative projects for a small island community on the frontlines of climate and social resilience. The event brought together a diverse mix of participants: local Rathlin residents and community leaders; academic facilitators from Queen’s University Belfast and Ulster University; 35 invited ‘design fellows’ and ‘ecosystem engineers’ (based across Ireland and the UK, from various backgrounds such as architecture, design, engineering, and marine biology). Rather than presenting a fixed project brief, the residential’s guiding philosophy was to allow ideas to emerge from the design fellows themselves. In practice, this resulted in the design fellows desiring extensive local input and context immersion preceding any design actions.

2.7.1. Process and Activities

At the pre-residential workshop, design fellows emphasised how important they felt it would be to encourage listening and learning during the residential. Instead of being assigned top-down, design teams self-organised around emergent themes that resonated with both participants and place. During reflection sessions, clusters of interest formed around topics such as wildlife, materials, water, building ruins, and regenerative tourism, which later became the ‘teams’ that crystallised. Individuals gravitated to the group that interested them, resulting in teams of varying size. Notably, the briefs were co-created in the moment: for instance, the materials team started broadly exploring local resources and honed in on the abundant wool from island sheep as an underutilised asset. At this stage, each team did not develop a brief; instead, they examined problems and opportunities on the island. This open brief-making is a hallmark of allogenic, community-led design, where the problems to be solved are defined through place, rather than predetermined by an external authority.

The immediate impact of the Field Operations proposals was difficult to quantify. Some island residents attended selected events and responded positively to specific ideas, such as a wool cooperative or ecological tourism concepts, while others remained wary, sceptical, or disengaged altogether. This scepticism cannot be dismissed as resistance to participation; rather, it must be understood within the more extended history of repeated consultations and externally led engagements on Rathlin, many of which have failed to deliver sustained outcomes. Prior one-off engagements (e.g., a 2018 charrette [21]) had limited legacy, contributing to consultation fatigue [22,23].

In light of this, it is essential to acknowledge the extraordinary effort made by the Future Island-Island team and design fellows to earn trust and build authentic dialogue. Early interactions took place not in formal settings, but in informal ones—most notably, the local pub (McCuaig’s, see Figure 4), where conversations were more reciprocal, casual-toned, and power dynamics less pronounced. Over time, some residents became willing to share stories, historical concerns, and critical reflections, which fellows treated not as data points but as material for more profound empathy and situated understanding. Participants reported that they gained “a deeper appreciation of the power of craft and storytelling in connecting to place,” having spent time weaving, foraging, and engaging in everyday rhythms with Rathlin residents.

The freedom to self-organise was empowering but came with challenges. Within each group sat ‘ecosystem engineers’ (invited to participate, see Appendix A for origins of terminology) and ‘design fellows’ (who had won their place through a design competition). Since no formal leaders or roles were assigned, the teams had to negotiate their own working structure. Some groups gelled quickly, especially those with around six members, finding it easier to share tasks, organise online meetings, and make decisions.

The role of ecosystem engineers warrants particular note, as it is distinct from conventional ‘experts’; these individuals acted more as critical friends or provocateurs, offering lateral input rather than direct leadership. Their role was to challenge assumptions, introduce new frames of reference, and subtly steer discussion without dominating it. In doing so, they helped mediate between bottom-up community insights and top-down design strategies—functioning, in effect, as facilitators of productive friction within the team dynamic.

In a few cases, one or two proactive individuals naturally assumed facilitation roles, guiding the discussions. However, larger groups (8+ people) struggled at times with ambiguous roles and direction, and with many voices and no clear hierarchy, discussions in those teams could loop without resolution. A degree of floundering was observed as they tried to decide on an approach. For example, one group of about 15 people passionately brainstormed dozens of ideas but struggled to converge on a single concept to develop, even after residential facilitators stepped in to help narrow their focus. This highlights a known insight from charrette practice, that inclusive collaboration needs gentle but purposeful steering. As the Project Management Institute charrette guide notes [21], effective facilitation is critical in sprints such as this. The facilitator’s role is to guide the process, keep discussions on track, and ensure that all participants can contribute effectively and meaningfully. In Field Operations, the organisers consciously took a back seat to let teams find their way, intervening only to ensure teams were meeting regularly, and checking in for observations and light feedback. With reflection, a more structured approach or shorter daily checkpoints might have prevented the loss of momentum and led to stronger outputs. However, the self-directed format did succeed in offering the flexibility to pursue what they felt was most relevant to Rathlin at a personal level.

2.7.2. Outputs and Outcomes

By the end of the residency, each team crafted a concept proposal loosely responding to their chosen theme, which they presented to the wider group. The nature of these outputs was generally exploratory and conceptual rather than fully detailed designs with continuity. For instance, the regenerative tourism team created a short film to document what it was like to visit the island as a tourist and then discussed potential questions we can ask ourselves when thinking of a regenerative strategy. The research emphasised the importance of establishing a symbiotic relationship between visitors and the local community, to ensure that tourism is not merely an extractive process but instead fosters mutual benefit, where both visitors and locals gain value. Within the materials team, outputs were not strictly defined from the start but emerged organically; for example, craft workshops became learning toolkits that not only helped people learn new skills but also acted as icebreakers for conversation starters between residents and design fellows. There was a strong emphasis on craft and hands-on activities that created cross-generational engagement, with both young and older participants actively involved (see Figure 5).

From a methodological standpoint, Field Operations embodied both the strengths and limitations of an allogenic, place-led co-design approach. It honoured resident voices, surfaced embedded knowledge, and created conditions for trust. Nevertheless, it also revealed the vulnerabilities of such a model: without a clearly articulated implementation mechanism or a strategy for transitioning from insight to action, the process risks stalling. This is a standard tension in community co-design globally, where collaborative processes generate meaningful relationships but struggle to maintain momentum or yield deliverables without structural follow-through [18]. By the final day, some participants expressed a desire for more crystallised and refined outputs, noting that prolonged early-stage exploration may have constrained their ability to converge toward actionable conclusions.

Ultimately, the legacy of Field Operations may reside less in immediate proposals and more in the groundwork it laid for future engagement. Its success should not be measured solely by design outputs, but by how it grappled with inherited scepticism, recalibrated power relations, and modelled ethical listening. These lessons point toward a growing international recognition that co-design in community settings must be relational, iterative, and historically aware, particularly when communities have experienced extractive or one-off engagements in the past [24,25].

2.7.3. Post Field Operations Reflection

Following Field Operations, our work package team came together to reflect on the process and outcomes. While the Rathlin residency had been valuable for relationship-building and deep contextual engagement, it also exposed fundamental limitations, particularly the slow pace and difficulty in progressing toward actionable proposals. Some participants and facilitators expressed frustration at the lack of momentum and clarity around outputs, and this collective reflection shaped our decision to develop DesignLink as a next step: a deliberately contrasting format that prioritised structure, pace, and tangible outcomes. Partnering with organisations made strategic sense, as unlike place-based community contexts, businesses and institutions often have more explicit mandates, decision-making structures, and the resources to act on new ideas. By shifting to an organisation-partnered design sprint, we aimed to test whether more predefined frameworks could unlock creativity more efficiently while still addressing real-world sustainability challenges with design thinking in line with the green transition.

2.8. DesignLink: An Organisation-Partnered Design Sprint

DesignLink was a three-day intensive design sprint (held in Belfast in QUB School of Architecture, June 2025—see Figure 6) that provided a stark counterpoint to Field Operations. As the next part of the Future Island-Island work package initiative, DesignLink’s goal was to generate innovative strategies for regional businesses and organisations, harnessing design thinking from early-stage creatives, all in line with the green transition. In contrast to Rathlin’s place-based, community-oriented process, DesignLink was company-focused and tightly structured, embodying a mostly autogenic approach where the framework was set in advance and design teams drove the creative process on a rapid timetable. The event brought together 23 early-stage creatives (graphic designers, architects, artists, and more), who were referred to as ‘design guilds’, and grouped in pairs, with one group having three members. Each group was paired with a local company or public-sector organisation that acted as their project partner (effectively, the ‘client’). Importantly, each company came with a predefined brief for the team to tackle, such as developing a tea franchise for Northern Ireland or rethinking street furniture for an outdoor museum. These briefs were curated by the DesignLink organisers in advance to ensure relevance to the themes of sustainability and community benefit.

2.8.1. Structured Sprint Format

Before the design sprint began, the organisers hosted a launch day the week prior. At this event, the design guilds had a chance to meet, the pairing structures were announced, and the briefs were assigned. This allowed design guilds the chance to familiarise themselves with each other and also to set the high energy tone expected for the design sprint. When the design sprint began and guilds arrived on Day 1, there was no ambiguity about what they would do or with whom—everything was predetermined: teams, partners, problems, schedule. After a kick-off introduction, teams met with their partnered organisations, which were mostly arranged through calls, but some also included site visits. This up-front structure enabled them to start immediately. The design process was then executed in a typical sprint fashion [21]: Day 1 focused on understanding and mapping the problem and initial ideation, Day 2 involved refining up to five concept strategies and creating a prototype or detailed proposal, and Day 3 finalised the project and included a presentation. Facilitators (from the universities) monitored each team’s progress, keeping them on track to meet interim milestones (e.g., by the end of Day 1, have at least three idea sketches; by the end of Day 2, decide on one concept to develop). The atmosphere throughout was high-energy and focused, where organisers deliberately fostered a fun, collaborative energy—they used icebreaker activities but realised it was important not to refer to these as ‘icebreakers’, as forced communication could lead to overthinking of what to say next. DesignLink supplied continuous refreshments and encouraged teams to fill their workspace with post-its and sketches to make the process visible and lively, with reminders for constant documentation through photographs. This positive environment had a psychological effect, with participants reporting feeling their “expectations were exceeded” during the sprint, as the clear structure freed them from worrying about process and allowed them to channel all their creativity into solutions. As one design guild put it, “It was full of interesting conversations, lots of laughter and thought-provoking work”. The DesignLink experience strongly reflected that: it gamified intense work, harnessing friendly pressure and team spirit to turbocharge design thinking and innovation.

Although the setup was more autogenic, stakeholder input was not ignored. Teams had direct access to their partner organisation’s representatives, and many teams conducted quick user research within the limited time. For example, guilds working on a brief for a rural transport app managed to conduct some questions about current users (via phone/video call) on Day 1 to gather insights, which they fed into their designs. Another team tasked with creating an upcycling solution for a pottery manufacturer took a rapid tour of the company’s small studio to see the clay waste materials firsthand. These touchpoints ensured that the designers’ rapid ideation was grounded in a real-world context, which was a nod to allogenic principles within the overall autogenic sprint. Notably, unlike Field Operations where the stakeholders (islanders and the island itself) were considered co-creators throughout, in DesignLink the stakeholders (company staff and consumers) played more of an informant or client role: they provided initial information and periodic feedback, but they generally let the design guilds take the lead in generating ideas—a difference that was subtle but important, meaning design guilds felt permission to be bold and had agency over the outputs. Indeed, participants noted that they were comfortable proposing wild or futuristic ideas to the companies because, with businesses, the people engaged are essentially there in a professional capacity, and it is their paid job. As the design guilds offer creative consultancy, they are free to push boundaries. With a community like Rathlin in Field Operations, it is more personal—it is their home, and so design fellows were more cautious. This observation encapsulates a theme we will revisit—design guilds in DesignLink did not hesitate to exercise their expertise and imagination, whereas in the Rathlin residency, design fellows and ecosystem engineers often self-censored or held back until guided by the island’s wishes.

2.8.2. Outputs and Immediate Impact

The deliverables from DesignLink were crystallised in advance and communicated to the design guilds on launch day. Each team produced a well-defined design strategy or proposal, which they showcased in a final presentation session attended by all partner organisations and other invited guests. For example, one team partnered with a local architecture firm that sought ideas on how to utilise their office outside of working hours. Design guilds presented schematics of the strategy, a mock timetable plan for community use, and even a small physical model demonstrating how the building can be used for local purposes. The partner organisation was enthusiastic and indicated they wanted to pursue further funding to pilot the idea. Another pair of guilds working with a regional museum developed a scalable, transferable design strategy for public furniture that could be used on-site, incorporating sustainable design and giving back to the local economy. They not only designed and completed the drawings for this but also storyboarded a marketing campaign for it. The reaction from partners was overwhelmingly positive; many were surprised at how much fresh thinking and tangible output had been achieved in such a short period. The design guilds were immediately asked if they would be willing to present this same idea back to the higher museums board. Several organisations expressed intent to implement or further develop the ideas, as one stakeholder from a design festival organisation commented, “We are looking forward to seeing what we can do with these ideas and how we can implement them” (quote from DesignLink final event). In at least two cases, the sprint outcomes directly led to follow-up actions: a design guild was invited to apply for a position at the company by submitting their CV, and a client also used a guild’s proposal as the basis for a successful funding application to further develop the concept. This kind of near-term implementation provides strong validation of the mostly autogenic sprint model, as it yielded strategies that were not merely imaginative but also actionable and closely aligned with stakeholder needs.

Equally important were the professional and educational gains for participants. By working closely with real clients under tight deadlines, the young designers sharpened skills in communication, rapid prototyping, and user-centred thinking. They also built networks: many left with new contacts in industry and a portfolio piece grounded in a real-world challenge. The sprint’s structure, which demanded decisiveness and creativity under pressure, seemed to amplify the designers’ confidence in their own process, which was unlike the Rathlin residential, where some designers felt unsure if they were ‘allowed’ to propose certain things without community sign-off. This contrasted with DesignLink, where the designers took ownership of the problems and ran with them. We can observe that Field Operations taught humility and listening, while DesignLink taught assertion and rapid problem-solving. Both are critical competencies for climate crisis innovation: we need designers who can empathise and engage, but who can also cut through complexity to propose strategies quickly, in keeping with the truth that we are in fact in a ‘crisis’.

In summary (see Figure 7), DesignLink demonstrated the power of a mostly autogenic approach: clear structure and a limited timeframe, combined with the skills of a motivated design team, can yield remarkably tangible outcomes in a short span. It showed that engaging stakeholders need not mean offering the pen; instead, stakeholders can set the directions and then allow designers to generate options, checking in at key points. Interestingly, while this format was nearly the opposite of Field Operations, it was not absolutist in excluding co-design elements, as the presence of partners and client feedback meant the designs were not created in an isolation bubble. This balance was perhaps why the design sprint was so successful: teams did not spend days finding the problem (that was refined and handed to them), so they could spend their time solving it creatively; yet, because the problems were real and stakeholders were watching, the strategic response remained relevant and grounded, but with big picture thinking. The experience of DesignLink thus offers a complementary perspective to Field Operations, setting the stage for a comparative discussion of themes in co-design practice.

3. Results

3.1. Themes and Comparative Insights

Examining Field Operations and DesignLink side-by-side reveals deeper insights about how co-design processes can vary, and further, how those variations impact outcomes. Key themes emerging from these case studies are discussed below, with a brief overview in

Table 2. Comparative Table: Field Operations and DesignLink.

Dimension	Field Operations	DesignLink
Context	Rural island community with deep place-based ties and resource sensitivities	Urban organisations with predefined agendas and professional framing
Duration	5-days (residential immersion, shared meals, evening workshops)	3-day sprint (time-boxed in professional settings)
Briefs	Emergent, co-created with community stakeholders	Predefined by organisers in collaboration with partner organisations
Team Structure	Self-organised, variable size; thematic groups (e.g., Birds & Materials)	Small, fixed pairs/trios, assigned to organisational briefs
Stakeholder Role	Co-creators, mainly embedded in daily life, residents engaged via workshops, making-led sessions, informal conversations	Clients/informants engaged at briefing and feedback moments, transactional and professional relationship
Outputs	Conceptual toolkits, exploratory, material experiments, narrative themes	Tangible, actionable strategies, clear proposals
Strengths	Deep trust, place-based sensitivity, reinforced local pride, personal impacts on young designers	Rapid results, professional validation, implementation traction (requests for board presentations, CV/job enquiries for further work)
Challenges	Slow pace, role ambiguity, lack of convergence, risk of ideas fading without follow up, some designers hesitant to act without local endorsement	Risk of overlooking nuance, limited community introspection, dependent on partner readiness and quality of briefs

. Detailed thematic findings are elaborated in Section 3.1.1, Section 3.1.2, Section 3.1.3 and Section 3.1.4 while evaluative comparison against structured indicators is synthesised in the Discussion (Section 4). In addition to the descriptive comparison in Table 2,

Table 3. Evaluation of Field Operations and DesignLink against predefined indicators.

Indicator Category	Field Operations	DesignLink
Approach	Format & aims: Open-ended, community-centred immersion. Alignment to objectives: Strong emphasis on trust building, heritage, and sustainability.	Format & aims: Structured, time-boxed sprint. Alignment to objectives: Clear focus on delivering actionable outputs for partners.
Process	Stakeholder engagement rhythm: Continuous, allogenic, multiple evening workshops, making-led activities, informal interactions across five days. Collaboration: Self-organised thematic groups encourage cross-pollination, though at times slowed by ambiguity.	Stakeholder engagement rhythm: Punctuated, autogenic, stakeholders engaged at briefing and review points, not continuously. Collaboration: Fixed small teams enabled efficient division of labour and fast convergence.
Outcomes	Tangible artefacts: Conceptual toolkits (e.g., archiving futures), material experiments (e.g., wool reuse), regenerative narratives. Relational outcomes: Deepened community trust, renewed creative confidence among young designers. Implementation potential: Limited follow up pathways, risk of ideas fading.	Tangible artefacts: Actionable strategies, prototypes, organisational proposals. Relational outcomes: Professional validation, network-building, requests for board-level presentations and CV/job follow ups. Implementation potential: High—clear ownership by partner organisations, immediate pathways to uptake.

evaluates both cases against the approach–process–outcome indicators defined in Section 2.6.

3.1.1. Team Organisation

Field Operations was structured around open-ended, self-organised teams that emerged through thematic affinity. Teams such as Ruins Group, Birds and Materials, and Cross Connections formed organically, drawing together designers, artisans, and local residents. This fluidity fostered a sense of ownership and supported bottom-up exploration but also created ambiguity in roles. Several participants described “analysis paralysis,” where decisions stalled without clear facilitation [11,26]. In other groups, such as Water Story (i), unexpected collaborations emerged, showing the potential for organic partnerships when structure is light.

In contrast, DesignLink operated with pre-structured teams of two or three participants, guided by predefined organisational briefs. Questionnaires indicate that contributors appreciated the reduced cognitive load and clear boundaries, with one respondent noting the sprint “pulled me in different directions in a short amount of time, but it was focused and achievable”. The fixed format channelled energy into concrete proposals without diluting creativity. Some participants also reported professional benefits, such as requests for follow-up presentations and CV exchanges with organisational partners.

3.1.2. Context of Engagement

One of the most striking differences between the two case studies was the ethical and psychological framing of stakeholder relationships. In Field Operations, participants were designing in a residential community with deep place-based ties and histories. The emotional weight of this context led some designers to defer decision making, feeling they lacked legitimacy to propose interventions. As observed in similar place-based co-design literature, designers in such contexts often self-censor or delay action to avoid appearing extractive or colonial [27,28]. Designers frequently deferred to local voices, at times hesitating to propose interventions without explicit endorsement. For example, Ruins Group participants reflected on their assumptions about islanders, later recognising their resilience and resourcefulness. This shift from scepticism to respect illustrates the weight of place-based engagement.

The residential setting also fostered immersive interactions: evening workshops drew strong community attendance, wool-making sessions built trust through hands-on activity, and informal conversations carried into meals and shared spaces. These settings blurred boundaries between designers and residents, emphasising relational learning.

Conversely, DesignLink operated in a professional, organisational setting, where stakeholders played more defined roles as clients or institutional representatives. The psychological contract was more transactional, enabling designers to take creative risks without the ethical hesitation seen on Rathlin. Participants felt freer to propose bold strategies, with one reflecting that “I would love my life if I got to do this for my day job”. The professional framing also heightened implementation focus, as ideas were clearly linked to partner agendas.

The transferability of co-design practices, therefore, hinges significantly on context. Community-centred co-design may require longer timelines, greater trust-building, and careful attention to local knowledge hierarchies [24]. Organisational co-design, by contrast, can benefit from faster cycles and clearer boundaries but risks becoming too problem-solving-oriented without wider reflection. Recognising these contextual dynamics is crucial for adapting co-design to other settings—whether urban neighbourhoods, educational institutions, or health systems.

3.1.3. Stakeholder Involvement

Field Operations adopted a deeply allogenic orientation, with community voices steering the framing of design questions. However, this led to tensions around how much authority designers retained in shaping solutions. In some cases, participants hesitated to move forward without explicit local endorsement—a challenge echoed in several precedent studies [7,29,30,31].

DesignLink, by contrast, was autogenic by design but strategically layered in moments of stakeholder input, particularly at the briefing and feedback stages. This produced a more efficient workflow while still rooting designs in stakeholder needs. The sprint exemplified a sequenced co-design model [32], where stakeholder engagement is frontloaded and punctuated, rather than continuous. As observed by Bason [33], this model helps avoid stakeholder fatigue and supports agile decision-making more effectively. Feedback indicates that this rhythm avoided stakeholder fatigue while still grounding proposals in user needs. For instance, partners requested further presentations to senior boards, and one participant reported being asked for a CV—evidence that engagement had practical traction beyond the sprint.

These differences underscore a broader contribution to the literature: the value of flexible stakeholder engagement rhythms. Rather than assuming that stakeholders must be present at every design step, the evidence suggests that timing matters more than frequency. Co-design becomes most productive when stakeholders shape the framing and direction, and designers are given space to develop options that are then iteratively reviewed and refined. This middle path contrasts with some normative claims in the participatory literature that deeper inclusion is often assumed preferable; instead, it supports a more pragmatic, outcome-oriented view.

3.1.4. Outcomes and Impact

The outcomes of the two formats diverged in type and trajectory:

• Field Operations generated concept-level proposals rich in contextual sensitivity. Teams produced toolkits (e.g., Archiving Futures tourism framework), material experiments (Birds and Materials resource reuse), and narratives of regenerative practice (Water Story discussions on sustainable tourism). These outputs sparked conversations and strengthened local pride but remained exploratory. Feedback loops were informal, and several participants expressed concern that promising ideas might “fade without follow-up.” Nonetheless, residential transcripts also revealed personal impacts: young designers described the week as “reinvigorating” their creative confidence after feeling lost post-graduation. The ideas generated served as conversation starters and provoked reflection but had limited actionable continuity. As noted by Vaajakallio and Mattelmäki [34], many co-design projects fail to bridge the gap between ideation and implementation due to a lack of handover mechanisms or resourcing.
• DesignLink delivered tangible, actionable strategies (see Figure 8 and Figure 9). Organisational partners received polished proposals within three days, and follow-up interest was strong: requests for presentations at higher boards, employment enquiries, and further discussions of prototypes. Participants reported professional validation, with one stating the sprint “exceeded expectations and showed me a new way of working.” The structured handover mechanisms provided clearer pathways for implementation compared to Field Operations.

Across both cases, outcomes extended beyond deliverables to include relational and motivational impacts. Field Operations deepened community bonds and highlighted local capacities, while DesignLink expanded professional networks and created momentum for organisational change. The implication here is that co-design’s effectiveness should be measured not just by inclusivity or creativity, but also by the continuity and ownership of outputs. This aligns with recent evaluations of co-design impact (e.g., [35]), which argue for a broader set of success criteria that includes relational, procedural, and implementation outcomes.

These differences in outcomes highlight the need for follow-through mechanisms across co-design phases; the comparative implications are unpacked further in Section 4.

Figure 8. Output example from DesignLink: Branding Proposal for Climate NI.

Figure 8. Output example from DesignLink: Branding Proposal for Climate NI.

Figure 9. Output example from DesignLink: Launch Proposal for Belfast Design Week.

Figure 9. Output example from DesignLink: Launch Proposal for Belfast Design Week.

3.2. Future Work

The contrast of Field Operations and DesignLink offers a critical lens through which to interrogate prevailing narratives in co-design. These case studies challenge the assumption that co-design should always prioritise extensive stakeholder control. While stakeholder engagement remains a central tenet of responsible and ethical design [27,35], our findings suggest that a dogmatic commitment to an allogenic process—where designers continuously defer to stakeholder consensus—can inhibit progress, frustrate participants, and ultimately diminish creative outcomes.

Conversely, an overly autocratic model that centres on only the designer’s initiative can risk overlooking contextual nuance and stakeholder legitimacy. This echoes concerns raised in the literature about extractive or tokenistic practices in design-led innovation [29,30]. The most effective approaches, as evidenced across our engagements, were those that actively toggled between these extremes—leveraging external inputs to shape direction, while allowing internal creative momentum to develop focused proposals.

This observation aligns with contemporary critiques of co-design’s conceptual ambiguity. As Luck [11] notes, “co-design” is frequently applied inconsistently across domains, with its core mechanisms poorly defined. Our contribution presents a grounded articulation of two distinct modes and proposes that co-design should not be understood as a static position on this spectrum, but rather as a temporal choreography across it.

In Field Operations, the initial deferral to community authority reflected a well-intentioned ethical stance. However, this led to delays and ambiguity, particularly where participants felt uncertain about their ‘right’ to design without extensive validation upon each decision. As other authors have cautioned, such ethical hesitancy can undermine the confidence and expertise designers bring to the process [1]. By contrast, DesignLink adopted a more assertive structure, yet retained the legitimacy of co-design practices by embedding user perspectives at key intervals. This enabled faster cycles of ideation along with higher implementation potential, while still aligning with user needs.

These findings contest a common normative stance in the co-design literature that more participation inherently equates to better outcomes [16]. Rather, our study highlights that the quality and timing of engagement are more important than the quantity. Moreover, both case studies suggest that in these examples, stakeholders appeared to expect designers to lead—not in opposition to participation, but as a complementary practice. The perceived legitimacy of the design team increases when they demonstrate initiative informed by listening, rather than passivity in pursuit of consensus.

As such, this work joins others (e.g., [26,32]) in advocating for strategically staged co-design models, wherein stakeholder input is sequenced across project phases: discovery (high allogenic input), synthesis (design-led), and validation (return to stakeholder review). This approach may allow both designer agency and stakeholder empowerment to coexist without conflict, which could be carried out with the following modalities:

• Hybrid Workshop Models

Based on our findings, we recommend hybrid co-design formats that deliberately sequence divergent and convergent modes of thinking. A potential structure might begin with a community-immersive residency (like Field Operations) to surface place-based priorities, followed by a focused design sprint (inspired by DesignLink) to prototype targeted interventions. This would help address the implementation gap observed in many participatory processes [34] by pairing relational groundwork with action-oriented delivery. This sequencing corresponds with the hybrid workflow elaborated in Discussion Section 4.6.

This approach aligns with calls in the literature for processual models of co-design that evolve in response to the problem space [36]. For instance, Steen et al. [32] advocate for a shift from ‘designing for’ to ‘designing with’ to ‘designing by’, suggesting co-design is a developmental arc rather than a fixed practice. This theory, combined with our observations, suggests that a hybrid model would operationalise that arc across phases.

2.

Transferability and Contextual Sensitivity

While this research was embedded in the socio-political context of Northern Ireland, the underlying insights are transferable. Many of the tensions observed—between designer assertiveness and stakeholder authority, and between open-endedness and productivity—are echoed across various domains, including urban planning, health services, and education [24,30]. However, contextual sensitivity remains vital, for example, design sprints with vulnerable populations may require more scaffolding, while corporate sprints may tolerate greater speed and risk [24]. We propose that future co-design work include an explicit context-audit phase, where project leads assess community readiness, decision-making structures, and cultural dynamics to inform the balance of allogenic and autogenic inputs.

3.

Future Island-Island as a Testbed

The Future Island-Island project itself offers a scalable research platform to test such hybrid models across contexts. As an AHRC-funded endeavour that convenes community members, academics, and practitioners to explore sustainable futures in Northern Ireland, it provides both the experimental space and civic legitimacy to iterate on co-design methodology. The project’s emphasis on green transition and local resilience presents a fertile testing ground for innovation in participatory methods, particularly those that combine deep engagement with actionable strategies.

Subsequent phases of Future Island-Island could integrate iterative co-design sprints with local authorities, education providers, and NGOs. This would enable the testing of multi-phase engagement, including initial immersion with communities to gather insights, a sprint-phase ideation process with designers and stakeholders, and post-sprint validation with implementation partners. Over time, this structure could generate a distributed ecosystem of interventions, each tailored to its micro-context but guided by a coherent methodological ethos.

4.

Contribution to Co-design Evaluation Frameworks

Lastly this study contributes to the emerging field of co-design evaluation by engaging with recent frameworks that emphasise systematic and multi-perspective assessment. Rather than focusing only on deliverables, evaluation should be embedded from the outset to manage expectations, amplify outcomes, support implementation, and motivate participation. Building on Wang et al.’s five-step framework [8], our dual case study illustrates how evaluation criteria can be negotiated with stakeholders, adapted across approach, process, and outcome, and communicated to sustain legitimacy. For example, while DesignLink demonstrated strong implementation follow-through, Field Operations excelled in building relational trust, underscoring how different co-design modes can be appraised through tailored yet systematic criteria. In doing so, our cases offer empirical grounding to the call for mixed-method, context-sensitive, and iterative evaluation practices when operating with co-design practices. The evaluation indicators trialled here (see Section 2.6) are revisited in Discussion Section 4.4.

4. Discussion

4.1. Strategic Hybridity: Why Oscillation Between Allogenic and Autogenic Matters

Co-design for the green transition benefited from strategic hybridity: intensively allogenic, place-based engagement built trust, surfaced contextual knowledge and sparked collaborations, while autogenic, sprint-based formats created pace, decision clarity, and immediate implementation traction. Together, the cases show that effectiveness improves when teams oscillate between externally led and designer-led moments across a project arc, rather than committing to a single stance.

Field Operations produced rich community interactions, new partnerships, and context-anchored concepts across multiple teams during the Rathlin residential week.

DesignLink, in contrast, generated rapid proposals aligned to partner briefs and prompted concrete next steps, including requests for board presentations and even employment enquiries.

4.2. What Worked Across Cases: Collaboration, Materials, Stewardship

• Collaborative design synergies
  Collaboration between designers, artisans, and stakeholders consistently amplified idea quality and relevance; group exchanges such as Ruins Group and Birds and Materials show how collective synthesis accelerated concept development and surfaced locally credible proposals.
• Toolkits and frameworks as portable assets
  Adaptable toolkits (Archiving Futures; Birds and Materials) provide a means to carry learning beyond single events, blending sustainability and aesthetics while remaining responsive to local heritage.
• Reimagining local resources and craft
  Field Operations repositioned local materials and craft as strategic assets (Cross Connections; Water Story), pairing traditional techniques with reclaimed resources to tell a sustainability narrative that strengthens community identity.
• Interdisciplinary design lab behaviour in situ
  The residential functioned de facto as an interdisciplinary design lab: co-located teams iterated, tested, and exchanged methods in real time, which increased idea volume and cross-pollination.
• Balancing heritage and modernity
  Teams negotiated heritage–modernity tensions (e.g., Ruins Group; Water Story) by pairing conservation sensibilities with contemporary proposals, reinforcing cultural meaning while enabling change.
• Ecological responsibility and feedback loops
  Both cases embedded ecological responsibility and iterative feedback. In Field Operations, workshops introduced doughnut-economy thinking and waste mapping; in DesignLink, sponsors iterated briefs in-session and sought immediate follow-ups.
• Mentorship and knowledge pathways
  Participants emphasised mentorship and skills transfer (e.g., wool practice, craft collaborations), highlighting the role of design programmes in preserving and evolving local know-how.

4.3. Where Formats Differ (and When to Switch)

The formats diverged in tempo, locus of agency, and type of outcome. Field Operations’ allogenic stance maximised context immersion and stakeholder voice but produced fewer immediate implementation steps; DesignLink’s autogenic stance increased decision velocity and implementation intent but required robust briefs and partner readiness.

Field Operations drew higher evening attendance, and hands-on activities generated more intimate trust. Outputs included material libraries, prototypes, and heritage mapping.

DesignLink created momentum with board presentation requests, CV/job enquiries, and strong participant reflections on the professional relevance of sprint work.

We infer a phase-dependent switch: begin allogenic to surface context and legitimacy; move autogenic for synthesis and scoping; return allogenic for testing with users; and repeat as needed.

4.4. Evaluating Effectiveness Against the Indicators

Using the predefined indicators (approach, process, outcomes), Field Operations scored strongly on stakeholder engagement rhythm (multiple evening events, diverse actors), collaboration, and relational outcomes (renewed motivation, new connections), while DesignLink scored strongly on decision pace, tangible deliverables aligned to briefs, and immediate follow-ups (presentation requests, employment interest).

Because contexts and durations differ, we stop short of causal claims; instead, we offer analytic generalisation: the pattern of strengths aligns with each format’s intended role.

4.5. Implications for Designer Agency

Operationalising designer agency as problem-framing, creative autonomy, and process stewardship clarifies how formats enable or limit agency. Field Operations constrained designer steering early to privilege context-setting and community authorship, while DesignLink widened agency to drive convergence and produce sponsor-ready artefacts. The oscillation across phases therefore calibrates agency rather than maximising or minimising it. This supports using autogenic and allogenic as analytical descriptors triangulated with canonical participation models, rather than as literal ecological categories.

4.6. Practical Recommendations: A Hybrid Workflow

The findings suggest a practical sequence that combines allogenic and autogenic elements into a hybrid workflow:

⚬

Context-audit (allogenic): short, resident-facing events and making-led encounters to surface tacit knowledge and map local materials and skills.

⚬

Framed sprint (autogenic): time-boxed synthesis with designers and partners to shape prototypes and implementation roadmaps.

⚬

Open-door review (allogenic): return concepts to residents and partners for critique and adjustment.

⚬

Iterative micro-pilots (autogenic–allogenic): low-risk trials (e.g., material reuse demonstrators) with embedded feedback loops.

⚬

Knowledge pathways and capacity: mentorship and community spaces to maintain skills, ownership, and momentum.

⚬

Each step is grounded in observed evidence: evening attendance and material activities in Field Operations, implementation requests in DesignLink, and calls for mentorship and community space across both cases.

4.7. Limitations and Transferability

The findings are based on two context-bounded cases with different durations and actor constellations, and indicators were populated from mixed qualitative and descriptive data within a limited follow-up window. We therefore claim analytic rather than statistical generalisability. Nonetheless, the pattern of when each format excels, and how switching calibrates agency, should be transferable to comparable place-based transitions and partner-led sprints.

5. Conclusions

This study has critically examined the balance between stakeholder-led (allogenic) and designer-led (autogenic) modes within co-design practices, drawing on two case studies: Field Operations and DesignLink, where each illuminated distinct strengths and limitations. Field Operations affirmed the value of deep stakeholder immersion and ethical reflexivity but exposed the risk of process stagnation when designer agency is excessively deferred. DesignLink demonstrated how structure, pace, and designer leadership can yield implementable outputs, yet also underscored the necessity of anchoring creative activity in contextual knowledge and stakeholder legitimacy.

Together, the cases underscore that co-design should not be understood as a singular method, but rather as a modular and adaptive process, capable of shifting between participatory and directive modes depending on phase, context, and wider objectives. This insight contributes to a growing recognition within the literature that co-design’s transformative potential lies in its strategic hybridity, not its ideological purity.

Future Island-Island, as a cross-sectoral and place-based research platform, provides fertile ground to further explore this hybridisation. By structuring engagements that oscillate between open-ended exploration and focused delivery, design teams may foster both community empowerment and tangible innovation. This approach not only aligns with best practices in participatory design but also meets the urgency of addressing complex challenges such as the climate’s green transition with both empathy and efficacy.

Ultimately, successful co-design strikes a balance between inclusion and action, responsiveness and leadership. This study advocates for a design practice that is neither deferential nor detached, but rather deliberately calibrated—a practice of informed authorship, where the designer listens deeply and then leads boldly.