Jobs We Make
Published in

Jobs We Make

Social Innovation through Deep Learning and Evaluation

Process Innovation

Social Innovation tackles ‘wicked’ problems of development. The Harvard Business Review (HBR) defines wicked problems as social or cultural problems that are difficult and complex to solve due to incomplete and contradictory knowledge systems and their interconnected nature with other problems of development. A social innovation approach establishes novel ways to use current resources, creates new networks, relationships and systems that untangle the web of wicked problems, bringing coherence to a complex world.

Catalyzing social innovation requires a combination of human centred and systemic approaches. Learning is key as it accommodates the evolving and unpredictable nature of complex systems, by ensuring the feedback is continuous and the systems and program is responsive to the contexts in which they function. There is a need, therefore, to design evaluations and learning systems that are mindful of contextual factors being flexible, iterative and responsive to changing dynamics within interventions.

DA is testing a complexity aware approach to monitoring, evaluation and learning with programs such as Work 4 Progress (W4P). W4P addresses the complex issue of livelihoods by understanding social and cultural barriers to employment faced by youth and women in rural parts of India. Its learning and evaluation system works to reiterate social innovation and ground-up action for enterprise development and job creation in rural India.

Adaptive Learning System

The learning system enables real-time feedback to inform implementation and practice in the short, medium and long term. It is based on techniques of dialogue, analysis and reflection, working towards the objectives of:

+ Building a robust learning system inclusive of system-wide perspectives, diverse community voices and interconnecting processes and functions

+ Integrating learning with design and operations to activate feedback loops for informed decision making and adaptive management

+ Enabling evidence-based knowledge-building to deepen impact and for acceleration

The diagram is a visual representation of the functions, roles and flows of the learning system developed as part of the Work 4 Progress India program by Development Alternatives team*

Interconnected Functions

As depicted in the diagram, the adaptive learning system implements feedback-centered mechanisms by connecting different functions of the program:

Data and Analysis:

The system implements qualitative and quantitative tools to collect and analyse data across the various stages to build a real-time evidence base.Qualitative tools include interview recording tool, network mapping, and ‘day in the life of’ which is an immersive tool based on principles of ethnography to engage with community members, and deepen empathy. Other tools like the community chat groups (on whatsapp) maintain a personal connection with communities, and facilitate learning and exchange among them as well. The data collected and generated is analysed through NVIVO which tracks changes in narratives in the program and relationships between nodes in the system. For example — if access to information is identified as a need in communities, solutions are co-developed or enhanced to address these needs.

Entrepreneurs part of the community chat groups have been sharing ideas on coping with the pandemic. Simple ways of maintaining a personal connect with communities have given us insights on their values and dreams.


The information and analysis gathered is regularly fed into Developmental Evaluation tools — maintaining an active loop between data, analysis and evaluation.

DE has been defined as an evaluation tool that can help social innovators develop social change initiatives in complex or uncertain environments. DA is implementing customised Developmental Evaluation (DE) tools to understand what is enabling or impeding desired change at different scales of the ecosystem, that is, the participant, prototype, platform, community and macro level. Findings from DE also compliment external evaluations conducted by third party agencies.

Recommendations from these evaluation tools and processes inform the programmatic decision-making (through the sounding board, management memos) and knowledge transference.

Tools for data collection, analysis and evaluation being implemented across social innovation processes

Adaptive Management and Implementation:

The system adapts to ground level complexities and dynamics by responding and re-designing processes and strategies for relevant solutions co-created with communities. The adaptive management function identifies effective strategies to adapt to emerging conditions.

It forms these strategies with other program members and partnerships — clients, advisors, regional partners, support service providers and communities. They are nonlinear in nature, given the need to respond to changing dynamics and circumstances.Through community centered feedback systems such as chat groups and community run radios, it also maintains a direct connect with communities and builds evidence-based knowledge and communication for active information flows and feedback across the system.

Evolution of the adaptive learning system

Knowledge Transference and Communication:

Lessons, evidence and information is synthesised into various knowledge and communication products and shared across platforms. A few examples of these products are toolkits on innovative prototypes for replication, analytical reports on diverse perspectives on social innovation, documentaries and human-interest stories.

View resources at Development Alternatives website

These are open access resources which are customised and disseminated through accessible modes through platforms at the community, regional, national and global levels. Through knowledge and communication we ensure our work is relevant at all scales, resonates with communities and meso-level stakeholders and is responsive to contexts and systems in which we operate.

The learning embedded within the wider process of social innovation therefore, rather than working toward clearly defined targets attempts to create new relationships between stakeholders, encourage the development of new technologies and new ways of using existing technology, and develop new networks and relationships between suppliers, producers, consumers and supporting organization. DA takes into account varied factors in the ecosystem (social, economic, political) and their interconnected nature. As a result, early signs are visible of benefits and success of adopting a complexity aware approach to learning and evaluation, integrated with implementation, rather than external to it. With time, such systems can enable cross-scale, inclusive platforms for learning and practice.

*Development Alternatives team: Shrashtant Patara, Kanika Verma, Prema Gera, Vrinda Chopra, Ankita Pant, Ankit Mudgal, Omkar Gupta, Debasis Ray, Saubhagya Raizada, Upma Singh, Stuti Sareen

Editorial Inputs: Kanika Verma

Authors: Vrinda Chopra, Stuti Sareen


Considerations for Monitoring, Evaluation and Learning in Social Innovation Platforms, JOSHUA FISHER, Columbia University



Get the Medium app

A button that says 'Download on the App Store', and if clicked it will lead you to the iOS App store
A button that says 'Get it on, Google Play', and if clicked it will lead you to the Google Play store


Perspectives and stories on creating dignified livelihoods for all. To contribute, mail us at or tag #JobsWeMake.