It took Bangladeshi NGO Uttaran numerous interactions and engagements with communities in vulnerable settings to realise its approach to working with them was flawed. Over the past few years, it has changed course, reimagining what community-led projects should embody in practice.

As a rights-based organisation, Uttaran began working with civil society leaders and community members in Southwest coastal Bangladesh since the early 1990s on numerous intersectional advocacy projects, including fighting for land rights and addressing waterlogging. Over the past few years, the grassroots non-governmental organisation (NGO) has reorganised its approach to facilitating community-led projects in water governance and climate adaptation: instead of NGO handholding, communities in vulnerable settings will plan, design, budget, and implement projects themselves through direct funding.

This renewed approach to working with communities stemmed from a realisation that cannot be traced back to a single project or programme. It was a collage of initiatives, encounters, dialogues, and interactions with communities over an extended timeframe that taught them the need for collaborative governance. An important principle of Locally Led Adaptation (LLA) approaches, collaborative governance is characterised by calls for negotiated roles and dialogue among multiple actors, ultimately shifting decision-making power to local communities – over what gets built, how, and by whose standards. Incidentally, LLA projects have sometimes genuinely worked towards an equitable shift in planning and decision-making to local stakeholders. However, critiques have argued that such projects often resuscitate the top-down dynamics that LLAs were designed to counter, including the reduced power of local actors in fund allocation; the risk of ‘local elite’ capture; the unaddressed structural inequalities; the dehumanising and depoliticising resilience-focused solutions; and agenda-setting by international actors or funders, among others.

And so the need for collaborative governance is also rooted in the colonial past, said Zahid Amin Shashoto, Uttaran’s Head of Climate Change and Water Governance programme. “For years, people in post-colonial societies forgot how to govern by themselves. With every self-governed, true LLA climate adaptation action, they are reclaiming that knowledge organically, and we need to give them that time and space.”

Arriving at that realisation, however, cost time, money, resources, and negotiations of ideas and expertise.

Re-learning to work with communities

Much of this learning traces back to the collective efforts of Uttaran, communities, and other entities to reintroduce a centuries-old nature-based sediment and tidal water management practice into the formal water governance system. The impoldering introduced by the Dutch in the 1960s not only devolved into a series of environmental fallouts (silted riverbeds, waterlogging, and saltwater intrusion into groundwater), but also social fallout, primarily displacing the traditional community-driven sediment and river management practice, which was later termed Tidal River Management (TRM). While community participation is essential to effective TRM implementation, implementing authorities often disregarded people’s role in local consultation.

TRM was not the only experience that forced Uttaran to rethink how it engaged with communities and the state. For instance, Uttaran began working on land rights in the mid-1990s, although a major success came around 2004, when the Bangladesh government allocated land to the landless communities they advocated for. It was only then that they realised their approach was ineffective. Until 2004, the NGO adopted a confrontational approach, prioritising protest and marches over dialogue and viewing the government understandably as an adversary. It almost always culminated in violence and threats, without commensurate results. “It pushed Uttaran to shift strategy to a ‘negotiated approach’, moving from confrontation to coordination with administrators, and focusing on dialogue and holding officials accountable while understanding why processes were slow. It produced better and more concrete outcomes, including land distribution to landless farmers,” said Zahid, adding it underscored the value of negotiating between different stakeholders to surmount an impasse.

Zahid recounted another anecdote that inspired Uttaran’s shift in approach towards communities. Unlike Dhaka, where the government subsidised water testing, distribution and pipeline maintenance, Southwest Bangladesh lacked testing, infrastructure, and subsidies. Seeing this as unjust, Uttaran spent $60,000 (own funds, Corporate Social Responsibility funds, and grants) to build an overhead tank and a groundwater pipeline system, skipping a feasibility study for lack of funds. By year two, the project proved unsustainable: households were too scattered for cost-effective testing and distribution, and without Dhaka-level subsidies, Uttaran could not sustain it. Zahid reflected that they had replicated a “foreign” urban model in a rural context, guided by emotion and values rather than vernacular knowledge. The failure was not about choosing the wrong infrastructure, but about choosing on communities’ behalf. Uttaran’s real fix, then, was to work closely with communities to rethink, redesign, and develop locally appropriate models rather than adopting a foreign model.

In 2022, Uttaran collaborated with the Dutch government’s Netherlands Enterprise Agency (RVO) for the Reversing the Flow (RtF) programme, which provides longer-term funding directly to communities in vulnerable settings. Much like Uttaran, whose approach evolved through a series of projects that revealed shortcomings in its earlier methods, RVO’s collaboration has also been shaped by lessons learned from its multiple interventions in Bangladesh.

Besides the RVO specifications that the projects focus on water and climate adaptation, LLA and resilience practices, communities could choose their leaders, identify their own priorities, and decide how to manage funds to implement projects that address their immediate socio-economic and environmental challenges at a landscape level (a landscape comprised one village or a cluster of villages beside a canal or a floodplain). Uttaran transferred the entire amount to each landscape’s bank accounts.

When communities relearn to govern

Like any international development projects, community-led projects, too, run into challenges and failures. However, instead of hand-holding community members to make the putatively ‘right’ decision, Uttaran’s objective was to step back and let the community encounter failure and apply their learnings to the next course of action. The goal was to remind communities in vulnerable settings of their agency and right to self-govern their socio-economic realities. “We only ensure all projects abide by four principles: pro-gender, pro-people with intersectional vulnerabilities, pro-environment, and pro-poor. People found these a reasonable ask and adopted them into their projects,” said Zahid.

Zahid illustrated his point with an example: In Southwest Bangladesh, communities agreed to build dykes to protect their homes and land from tidal surges. The core group of the village, which leads all projects, set a timeline, earmarked $8,000, and began the work during the monsoon. But a heavy shower caused part of the dyke to collapse, sparking conflict within the community. “We anticipated this, but we did not intervene. It was their engagement, and a failure that they had to recognise and resolve,” said Zahid. The main committee decided that they would not handle the technical work but would only facilitate decisions. If it is a dyke, only people with experience living near or working on dykes will be part of the implementation team. “The initial approach had overlooked the valuable insights of those with lived experience. They realised the importance of harnessing the knowledge and experience that already existed within the community. In a sense, they were relearning and restructuring their governance,” Zahid reflected.

However, communities governing themselves do not always guarantee ‘fair’ governance on the ground, said Zahid. It threatens the equal distribution of decision-making power among the community members, especially in heterogeneous communities – a mix of elite, poor, landless, and caste groups – where decisions tend to favour the elite. “For instance, a seemingly community-led decision to spend around $14,000 on a road turned out to be influenced by two powerful shrimp farmers, not the villagers’ actual needs. It showed how participatory decisions can mask elite capture, and how easily that goes unnoticed without someone on the ground willing to dig into who really benefits,” he said, explaining the pitfalls of community-led initiatives in highly unequal societies.

However, Zahid cautioned against writing off these challenges as failures, and justifying not allocating funds and shifting decision-making power to community-driven projects. “If NGOs found no success, it is often recorded as ‘a good learning opportunity’. Contrarily, community failures are often written off as ‘a failed approach’, risking an abrupt severance of funding,” he noted. Instead, he encouraged international development agencies and practitioners working on such projects to observe and learn how communities make decisions, implement plans, fail, learn, unlearn, and redo when given the freedom, space, and resources – “particularly because once donors, INGOs and grassroots NGOs leave eventually, it will be the communities who have to work alongside their government, and lead and sustain the adaptation efforts,” he reminded.

Addressing such ground-level challenges, he learnt, also demanded trust, time, and patience as investment. When residents are given ownership of funds, planning, budgeting and implementation, it may seem like a waste of resources – only when viewed from a short-term perspective. “When considered over a longer timeframe, we can observe how communities have broken away from NGO-designed development projects. They build their own assets, trade, understand their own capabilities, and reactivate their power and agency within societies. This will take five to seven years, and that is when you expect resilience, not within a year,” said Zahid.

Uttaran’s Failing Forward story is a reminder of what community-driven or people-led projects should look like in practice; not steered by NGOs or funders towards their own interests and understandings. Programmes built on collaborative governance – providing longer-term funds with a hands-off approach and interdependent partnership – are needed experiments in a sector that stubbornly continues the systematic top-down imposition of Western standards and, in the process, invisibilises other practices and reinforces societal hierarchies. If institutional frameworks such as the Logical Framework or numerical targets are the only focus, the top-down machinery will stymie the exercise of locally led adaptation plans, and continue to fail. As Zahid said, “Giving people back the governance systems is what failing forward perhaps looks like.”

In this Failing Forward? episode, Cristal Ange, General Director of Fundación Herencia Ambiental Caribe (Caribbean Environmental Heritage Foundation), reflects on how her understanding of community participation evolved from routine meetings to something far more personal for those participating. Drawing on a jaguar conservation project in Colombia’s Sierra Nevada de Santa Marta, she challenges how scientists, development practitioners and funders define participation, and how ‘what looks like success’ may, on closer inspection, be something else entirely.

In 2008, the National Parks System of Colombia approached Fundación Herencia Ambiental Caribe (Caribbean Environmental Heritage Foundation) to help them assess the jaguar population and plan conservation efforts in the Sierra Nevada de Santa Marta, a coastal mountain range in northern Colombia. Jaguars are a good indicator of a healthy ecosystem, as the wild cats favour dense forest, river, and wetland habitats, keeping a check on prey species and maintaining the ecosystem balance. When local communities are involved in jaguar conservation, it essentially helps protect the forests and waterways they rely on. Accordingly, the project engaged local communities and the indigenous group, the Arhuaco, to understand the jaguars’ path in the coastal mountains.

“It was an open and participatory project, where the Arhuaco people would gather scientific evidence about jaguar movement and their path. We travelled to different villages, organised meetings with the indigenous residents, and showed portraits of jaguars to enquire how they referred to jaguars and the sites where they were spotted,” recounted biopsychologist Cristal Ange, the General Director of the foundation, which focuses on sustainable water management and biodiversity protection by partnering with indigenous, Afro-descendant and peasant communities in Colombia.

Beyond gleaning information, the team trained the indigenous people to use camera traps and the methodologies to learn about jaguars and their trails. They also organised capacity workshops to help the indigenous people understand what they were doing, and how that information would be used.

After a year, the team closed the first phase of the project. “We found the jaguar sites, as well as collected pictures and magical stories to share with the world. We also understood how to cooperate with local communities. It was a success,” Cristal remarked.

“Or, was it really successful?”, she quickly paused to consider.

Cristal attributed her reservations about calling the project a success to the consultative and technocratic approach to community participation the team had adopted throughout. “We were too focused on what the science told us, and not enough on ancestral knowledge and ways of understanding the territory held by the communities,” she acknowledged.

Displacing knowledge pluralism

In the post-war era, community participation was adopted as an essential ‘tool’ for development projects in the Global South; as an antidote to the repeated failure to meet the project objectives and outcome; and to counter the traditional top-down strategies that failed to include the people’s voices.

Community participation became a precondition to funding. Projects designed participation approaches around scientific and technical knowledge frameworks, donor mandates, and a pre-determined set of criteria, while (paradoxically) expecting new knowledge to emerge from this renewed focus. And while the project runs, residents support it, attend meetings, and some may temporarily find a job. They are trained and armed with Western and Eurocentric rules and guidelines to approach the very territories they inhabit and interact with every day. The project led by Ange’s foundation to scientifically establish the jaguar path is one such case.

Sample this: An indigenous community member, familiar with jaguar territory and movement in Sierra Nevada, had predicted the animal’s presence before placing the camera traps provided by the team. He knew, without checking, that his camera had captured jaguars. The request for photographic evidence to trace the jaguar’s path sparked his curiosity, as it pointed toward forms of data and ecological information that extended beyond the traditional understanding of jaguar movements.

“Yet, our attention remained centred on the project’s technical requirements and the funders’ interest in ecological data about the species, rather than on the broader ancestral knowledge that had informed his observations,” she recounted

This serves as a powerful empirical illustration of unintentionally reducing the participant to a data collector instead of a knowledge holder. “Traditional Western means of participation incorporate people into how we, as scientists, think jaguars should be conserved. But what do we know about what is the right way for them?” Cristal reflected.

In its noblest vision, community participation is designed to promote dialogue, interaction, learning, and co-production of knowledge between development agents and the community members. In practice, though, community members are often consulted after the project is designed; they are excluded in the implementation, monitoring and evaluation stages; or are approached to resolve a particular information problem to achieve the predetermined project goal.

As a result, community participation has often unfolded within institutional and scientific frameworks, which open important spaces for collaboration between researchers and local knowledge holders, but at the same time, continue to be shaped largely by Western approaches to how knowledge should be validated, interpreted and incorporated into development projects. It systematically devalues and displaces indigenous people from their own everyday reality or their ways of knowing. Such approaches have reinforced themselves as a tool to legitimise Western, neocolonial projects rather than a veritable power-sharing and knowledge co-production procedure.

Paying attention to this contradiction entails self-reflection: did ‘new knowledge’ eventually emerge? Was ‘power’ distributed unidirectionally? Did it eventually devolve into the very ‘subject-object relation’ that community participation set out to eliminate?

Understanding the ‘other’ knowledge

Cristal and her team shifted their base closer to Santa Marta to spend a considerable time going into the jaguar territories with the tribal elders — not to organise workshops, but to sit down with the people, observe their daily activities, and understand their reality, rituals and knowledge, with humility. Now into the second phase of the jaguar conservation project, the team focuses on what matters most to the indigenous people: conserving the jaguar’s spiritual energy, which they believe is a living force, embodied in natural elements, such as a rock or sacred sites. “There is no scientific explanation, but it is the elders’ way of knowing,” Cristal challenged.

Their guiding principle: “If we want to conserve jaguars, how can people support science, and also how can we, as scientists, development practitioners and funders, support and learn about the local ways of conserving the animals.”

A rock identified during the project, for instance, illustrated this difference in perspectives. From an ecological standpoint, the site did not play a measurable role in jaguar conservation. Yet for the indigenous communities, conserving the rock was essential. It was understood as a sacred jaguar site where the animal’s spiritual energy resides and the broader balance of the territory is maintained. The project thus began to recognise that conservation is not only about protecting habitats and species populations, but also about sustaining the cultural and spiritual relationships that shape how communities coexist with the jaguar. “Our goal was not merely to acknowledge indigenous practices, but to validate their knowledge and embed it at the heart of the project,” she emphasised.

Success for whom?

Understanding failures entails a growing self-awareness of how practitioners and scientists can mistake failures for success. In the traditional methods of implementing development projects, participation includes organising workshops, capacity strengthening sessions, and handing down scientific ways of knowing and working with the local community. The latter’s ‘contract’ as participants often ends at this stage, with no active role in the evaluation and monitoring phases of the project cycle. When zooming in on this aspect, the project is a success – but for whom?

To the donors, scientists, and the NGO, the project was a success. When considering the project more holistically — including whether participation translated into changes that mattered in the daily realities of the community — the project may also reveal important limitations in what it was ultimately able to achieve for the people involved.

For such a feedback system to be practical and functional, one needs to go back to the drawing board to assess whether the project makes sense to the local community or the participants. As Cristal, too, reflects, if a project cannot explain how it ties into the people’s local persuasions, it will not work in the long run.  It may continue to operate and generate scientific data considered relevant for institutional and conservation decision-making, but if that knowledge does not also resonate with the priorities, relationships, and ways of deciding within the local communities themselves, its impact may remain limited once the funding cycle ends.

Another sticking point of projects that work in indigenous territories and with their people is funding. Supporting both ways of knowing demands a budget for what funders think should be done, but also for local ways of doing things. However, funding with the correct attitude towards local knowledge and beliefs, too, is critical. “A rock, for instance, may not have added to the ecological value of the jaguar corridor, but it added to the personal value of the project,” she challenged the rampant compartmentalisation among development agencies of what counts and does not count as knowledge.

For Cristal, the project’s most important lesson was not that science had failed, but that scientific frameworks alone were insufficient to understand the full complexity of coexistence in the territory. Being honest about those limitations opened the possibility of recognising other forms of knowledge that had long sustained relationships between people, jaguars, and the forest. In many ways, these approaches were already working on the ground — not necessarily through scientific indicators, but through rituals, practices, and sacred sites that had enabled coexistence across generations. The question is no longer whether these other ways of knowing exist or matter, but whether conservation institutions and funders are willing to invest in them on their own terms.

As the conversation drew to a close, she considered how people’s participation in knowledge co-production ought to be: “Development projects are not only important for the Western academic sector or NGOs. It belongs equally to the people. They may not engage with the project through the same technical or scientific frameworks in which the plan is designed, but they understand its importance and how it complements their own ways of understanding and relating to the territory.”

“I have not failed. I’ve just found 10,000 ways that won’t work.” Thomas Edison’s words could be the motto for this edition of the Waterproof podcast, where guests share what happened when water projects didn’t go quite as planned – and what the sector learned from it.

This article is based on the ‘Failing Forward’ episode of Waterproof, a podcast by Partners for Water.

A quiz show about menstruation might seem like an unusual way to start a conversation about water and sanitation failures. But that is exactly what Esther Shaylor, Innovation Manager at UNICEF and part of the WASH Failure Movement, did – using humour to break down the barriers that keep professionals from openly discussing what goes wrong.

In the Waterproof podcast episode ‘Failing Forward’ Tracy Metz speaks with four guests who have all, in their own way, learned that progress in the water sector often comes through trial, error and honest reflection.

Talking about failure openly

Esther is the driving force behind the WASH Failure Movement. She has spent years encouraging colleagues to speak openly about projects that did not go as planned. “We really want to talk about failure,” she says. “We want to openly and freely share it so that we can learn from it.”

That openness has surfaced some striking stories. In Afghanistan, pipes were laid to bring water directly to women’s doorsteps – seemingly a clear improvement. But for many women, the daily walk to collect water was also their only opportunity to leave the house and meet others. The result: some damaged the pipes themselves so that water collection, and the social contact that came with it, could continue.

Then there is the story of the jerrycan. UNICEF set out to design and manufacture an improved jerrycan for emergency settings, investing heavily in research and a custom mould. When it was finally tested with users, however, most still preferred the original product. “This is something I have learned repeatedly,” Shaylor reflects. “We need to define what success looks like before we start.”

When good intentions meet a different context

For Zahid Amin Shashoto of Uttaran, a grassroots organisation in southwest coastal Bangladesh, failure is often less about technology and more about context. In the 1960s, Dutch-style polders were introduced to support intensive farming. But unlike in the Netherlands, water that entered these polders could not easily get out, leaving the land waterlogged for months at a time.

“My community is primarily farmers,” Shashoto explains. “And if farmers don’t have farmland – if it’s underwater – they go out.” Many farmers leased their flooded land to large-scale shrimp farmers, introducing saltwater onto land that had supported rice cultivation for generations.

“Serious mistakes were made in thinking that Bangladesh looks very much like the Netherlands, which it doesn’t,” Shashoto says. “The technology itself isn’t bad. But introducing the technology without considering the context is a mistake.” A related lesson came from a project that provided arsenic and salt filtration systems for drinking water: highly effective, but too costly for communities reliant on government subsidies to maintain once the project ended.

Uttaran’s response has been to shift towards what Shashoto calls a “negotiated approach”: channelling funding directly to communities so they can address water, education and land issues together, as a collective.

A safe space to fail

At the Green Village, a field lab at TU Delft, failure is built into the design. “For us, it’s interesting to look at things that are new, that are not yet applied or not yet validated,” explains Lindsey Schwidder, programme manager for climate adaptation. “And here, they are also allowed to fail.”

One entrepreneur tested a pavement substructure made from recycled glass waste, designed to store rainwater underground. The product worked, but because it was officially classified as ‘waste’, every delivery required a separate environmental permit. The barrier was not the technology, but the paperwork. Eventually, the material was reclassified as a building material, resolving the issue.

Other experiments reveal similar friction. Lighter-coloured pavements could reduce summer temperatures by several degrees, but municipal style guides often prescribe specific colours, making adoption difficult. And for green parking spaces designed to let rainwater filter into the ground, it is not always clear which municipal department is responsible for maintenance. As Schwidder puts it, making climate adaptation work requires bringing legal experts, financial teams, contractors and maintenance departments to the table, not just engineers.

Failure as part of diplomacy

Even at the highest levels of international water diplomacy, things do not always go to plan. Former Water Envoy Henk Ovink, who now leads the Global Commission on the Economics of Water, recalls preparations for the 2023 UN Water Conference with Tajikistan. Early on, communication stalled: “There was a shopping list from the Netherlands on what needed to change. And on Tajikistan’s side, there was the question: why do we need to change course?”

By investing time in understanding each other’s perspectives, the relationship recovered and strengthened. “We made the mistake of being too preoccupied with our own ideas,” Ovink reflects. “But we learned fast, and it really paid off.”

For Ovink, this is the essence of failing forward: “A lot of things go wrong before things really go right. Learning is exactly what failure is all about.”

As Edison put it, failure is just another way of finding out what does not work. Across the water sector – from UNICEF’s jerrycans to Bangladesh’s polders, from Delft’s experimental pavements to UN water diplomacy – the willingness to examine what went wrong, and why, is what turns setbacks into progress.