In complex, collaborative development initiatives – where outcomes emerge over time, and values may differ – evaluation needs to do more than measure results. This post outlines why and shares five grounded patterns from complexity-aware monitoring, evaluation and learning (CAMEL). Finally it reflects on what this approach asks of evaluators in practice.
I’ve worked in long-term programmes where the most meaningful shifts weren’t always the easiest to measure. Changes in how partners related, how strategies evolved, or how systems aligned often made the real difference – but didn’t show up neatly in a dashboard.
In more conventional projects, evaluation focuses on fixed plans, short-term deliverables, and defined outcomes. But system change efforts unfold differently. They span longer timeframes, involve many actors, and aim to influence relationships, behaviours, and institutions. Outcomes are often emergent, and success looks different depending on who you ask.
This is where complexity-aware monitoring, evaluation and learning (CAMEL) comes in. It supports reflection, learning, and strategy – not by applying fixed criteria, but by helping people make sense of change as it happens.
And yet, when it comes time to evaluate, the pressure is often on to “prove impact.” Funders want targets. Teams want time to reflect. Outcomes are still unfolding, and the real questions aren’t just about what has changed, but how, why, and what it means.
This tension is common in areas like climate adaptation, food systems, freshwater management, and land use. These are areas where values diverge, and linear models of change fall short.
Many programmes in these fields operate in complex systems. Here, cause and effect are rarely straightforward, perspectives on value vary, and change emerges over time. Conventional approaches, especially those tied to predefined plans, can struggle to stay useful.
CAMEL offers a more adaptive alternative. It draws on systems thinking, participatory learning, and reflective practice to support real-time strategy and adjustment. When answers are uncertain and outcomes evolve, it helps teams pause, ask better questions, and notice what’s shifting.
While conventional methods remain useful for tracking defined outputs, CAMEL complements them with tools that support learning in dynamic settings. It also reframes accountability – focusing less on control, and more on explanation, adaptation, and transparency.
These reflections draw on long-term experience in collaborative evaluation and systems-focused programmes, together with learning from the broader field. In particular, it aligns with contributions from Herz et al. (2021), Rosenberg et al. (2025), and the Social Finance team (2023), all of which are highlighted on the Learning for Sustainability resource page on complexity-aware MEL.
Five patterns in complexity-aware MEL
Drawing on this research and practice, I’ve come to recognise five recurring patterns that characterise complexity-aware MEL. They’re not steps to follow but overlapping ways of working that reinforce each other.
1. Begin with grounding and orientation
Evaluation doesn’t start with metrics, it starts with people. In complex initiatives, that means building relationships, understanding context, and exploring how different groups see the purpose and direction of the work. This includes acknowledging history, inequity, and diverse values, and creating space to discuss what matters.
Developing a theory of change is part of this, not as a fixed blueprint, but as a shared orientation. While initial versions may be drafted by a lead agency or team, their real value lies in supporting ongoing dialogue and alignment.
Rather than focusing on fixed outcomes, CAMEL encourages attention to trajectories of interest – directions grounded in shared values. These help track shifts in relationships, systems, and practices over time.
This early orientation also informs later monitoring and evaluation. By surfacing key dynamics and uncertainties, it helps shape flexible indicators and supports adaptive management.
2. Make learning central, not an afterthought
In complex settings, learning isn’t an add-on, it’s how programmes adapt. That means making space for reflection throughout an initiative, not just at the end. This might include ‘pause-and-reflect’ sessions, peer learning loops, or structured debriefs. The aim isn’t to capture everything, but to prompt timely insight and course correction.
Theory of change work becomes valuable here, not as a static model, but as a tool for testing assumptions and adjusting direction. MEL becomes a support for learning, not just accountability.
Finding time for reflection isn’t always easy. Delivery pressures can crowd it out, and monitoring may feel like a tick box exercise. But without it, teams risk losing sight of where they’re heading, or what they’re learning.
The aim is to support social learning which extends beyond the programme team. In many initiatives, insights are shared across sites, with partners, or through tools that prompt wider reflection. Evaluators can support this by holding space for inquiry and helping programme teams and partners to use insights as they go.
3. Hold space for multiple perspectives of value
In complex initiatives, there’s rarely one view of what counts. Funders may prioritise measurable outcomes. Communities might focus on relationships or cultural continuity. Agencies and other stakeholders bring their own values. Rather than forcing consensus, complexity-aware evaluation holds space for difference. Collapsing everything into a single frame risks flattening what matters most.
Appreciating diverse perspectives means actively seeking different kinds of contribution. This might involve interviews, co-developing rubrics, or using methods like Outcome Harvesting or Most Significant Change. These help surface relational, ecological, cultural, and economic contributions – especially those linked to legitimacy, policy influence, or long-term system health.
Intermediate outcomes – like improved data practices, shifts in governance, or shared infrastructure – often signal the potential for deeper change. They create the enabling conditions for long-term impact, even when final outcomes remain uncertain.
Evaluators can support this broader view by asking: Whose definitions of success are shaping the story? What forms of value are visible, and what might be missing?
4. Support systemic co-design and adaptive management
Complex initiatives rarely follow a straight line. Evaluation can help teams zoom in and out – across timeframes and scales – to notice both details and the bigger picture. This can surface unseen relationships, feedback loops, leverage points, and blind spots.
This is especially important in initiatives that span multiple systems – like climate adaptation, freshwater management, or food systems – where progress depends on working across sectors and institutions. When teams are deep in delivery, shifting dynamics can be hard to see. Evaluation helps reveal what’s changing, how people are responding, and where new opportunities may lie.
CAMEL approaches work well alongside systemic co-design. These collaborative, multi-stakeholder processes often unfold over extended periods and involve multiple layers of decision-making. Evaluation helps support this by aligning learning with direction, connecting what’s emerging on the ground with broader system goals.
Evaluators can support adaptive management by helping teams trace shifts and surface the thinking behind them. In complex systems, action evolves with learning. Evaluation helps legitimise change, prompt reflection, and support reorientation.
5. Design for reflection and use
In complexity-aware MEL, use isn’t a final step, it’s designed from the start. Insights land best when they’re co-created, timely, and relational – not delivered as verdicts after the fact. That’s why participatory approaches are common here – not as an add-on, but as part of a different stance. Rather than reporting from the outside, evaluators co-create findings with others.
Interpreting insights together, co-authoring reports, and co-designing simple reflection tools all help teams internalise learning. These shared processes provide valuable moments of reflection, not just for the team and partners, but for the evaluator too. The goal is less about delivering conclusions, and more about enabling real-time inquiry, insight, and adjustment.
Tone and timing matter. Evaluation often holds up a mirror.In sensitive settings, how something is said or how findings are shared can shape how they’re received. Sharing insights in ways that support relationships without diluting what matters helps teams reflect, reframe, and move forward.
Designing for use also means leaving something behind – not just a document, but shared language, reflective habits, and strengthened capability. In some initiatives, final reports became prompts for internal dialogue or partner workshops. That signals a learning orientation that is active, embedded, and relational.
What this asks of evaluators
Working in complexity-aware ways changes how evaluators show up. It’s not just about using tools differently, it’s also about adopting a different stance. Rather than arriving as external judges, evaluators using these approaches are often more effective when they step in as learning partners. It means creating space for reflection and helping people ask the questions that matter most.
In complex initiatives, causality is often unclear, outcomes emerge slowly, and people value different things. Evaluation becomes less about delivering answers, and more about helping people make sense of where they’ve been, where they are, and where they might go next. In complex situations, a well-timed question often opens more possibilities than jumping to answers.
Often, this means working across multiple accountabilities – delivery teams, strategic leaders, funders – each with their own priorities. Part of the role is helping these groups stay connected to purpose without losing sight of complexity.
Evaluators can also help create space for shared sense-making. That might mean light-touch tools, inviting diverse perspectives, or naming tensions in ways that support learning. In my experience, the most meaningful insights often emerge in these quiet moments, not through formal reporting alone.
Scaling is another part of making evaluation useful over time. If you are working with change initiatives that need to travel across places or partners, this short introduction to rethinking scaling in complex settings may be helpful.
None of this means giving up on independence. But in complexity, independence is held differently. It’s not about distance, it’s about perspective. Evaluators working in this way stay grounded in purpose, ask questions that surface insight, and help others sharpen their own judgement. The rigour comes from reflection, shared inquiry, and staying close to what matters.
Closing thoughts
Complexity-aware evaluation isn’t just a method, it’s a way of showing up. At its heart, CAMEL supports people and organisations to reflect, adapt, and learn their way into change. It asks us to see evaluation not as a moment of judgement, but as part of the longer arc of system stewardship.
That means designing CAMEL not just for short-term accountability, but to support longer-term coherence and connection. Done well, it strengthens shared intent and builds the relational infrastructure – trust, capability, and shared language – that enables work to continue through change.
CAMEL also recognises that value takes many forms. Environmental, cultural, social, and economic contributions all matter—at different scales and from different perspectives. This approach holds those views in conversation, without rushing to resolution.
And it reminds us that evaluation doesn’t just monitor progress, it shapes it. The way we design and implement CAMEL influences who gets to learn, what gets surfaced, and how decisions evolve. That’s why embedding the approach in planning, governance, and relationships matters. It’s not just technical, it’s strategic and ethical.
Yes, this work takes time. But sustainable change doesn’t happen on short cycles. Complexity-aware approaches help us to stay with the work long enough to learn, adjust, and make progress that lasts.
This post also sits within a broader reflection, as Learning for Sustainability marks 20 years online in 2025. Many of the ideas explored here draw on that work – supporting participatory processes, systems thinking, evaluation, and collaboration in diverse settings.
[* Image: AungMyo / Adobe Stock]
You’ll find additional tools and resources on the site, including pages on complexity-aware MEL, Planning, monitoring & evaluation, Theory of change, and Logic modeling, If this post was useful, feel free to share it with others. You can also sign up for occasional site updates to hear about new tools, guides, and key resources.
