The best ideas emerge when very different perspectives meet – Frans Johansson
…and better impact emerges when different approaches are combined. In the field of education as well, it takes all kinds to move the needle – whether we are working at scale, or in a small number of schools, or in teacher development, or in technology. And often, figuring out where and how we want to contribute is a journey of evolution. Our journey was no different – from an organization working with a small number of schools to now working across multiple states; it has been a journey of evolution in our thinking and in our actions. We recognized that in order to move closer to the vision we have imaged, we’d need to work closely with state leadership and enable large-scale improvement.
There were multiple articles, lectures, and videos on why we should think about scaling, successful cases, pros and cons – of working at scale – but ultimately, it has to be a (personal) decision of the organization. Once the decision was made, though, it was an even bigger roller coaster ride! Our team had been used to a certain way of working, which now had to change significantly. There were questions we asked ourselves every day – and over time, it has led to a (fairly) shared understanding of what working at scale means to us as an organization. For those of you who have made a decision to work at scale but are figuring out the kinks, and aligning your team, these questions may come up for you too!
How do we do the same thing with 10,000 people, which we were doing with 10?
Short answer: We don’t. “Designing for scale” is perhaps an incomplete term. What we really need to think about is “Designing for impact at scale”. Trying to do the same with 10,000 people, which we did with 10, is not scale, linear growth, or replication. What we needed to think about was – what are the areas of impact we want to see at scale, and how might we achieve those? For example, when we worked with 10 schools, a key intervention was conducting whole school evaluations so that improvement strategies could be more relevant and need-based. This included our team members going to each school, administering various tools like surveys, interviews, FGDs, and observations to gain a holistic picture of the school. When working at scale, doing the same activities were not possible – but it was possible to change our strategies to achieve the same results (of getting an understanding of key areas of improvement of schools). So our strategies changed – now instead of us visiting schools, we enabled school leaders to assess their own schools (which included re-designing the assessment so that leaders would be able to understand and apply the tool), we worked with partners to enable the assessment on technology, and we leveraged forums like YouTube Lives and courses on DIKSHA to guide leaders on the assessment. What we learnt from this was:
1. Design keeping the goal in the center, and not the activity/strategy.
2. Think about what would work at scale, rather than scaling what works.
3. Leverage partnerships
Things are in our control when working with x schools. They won’t be with 100x.
This question probed us to really reflect on our values – was it important for us to be able to “control” what happens and how it happens, or did we also want to empower people to solve their own challenges, and lead their own improvement? Things don’t need to be in our control. Will things go exactly as we planned? Probably not. Will the impact be different? Likely. Will it empower more problem solvers? Absolutely. In Uttar Pradesh for example, there is a state-wide focus on supporting children to attain foundational literacy and numeracy skills under the Nipun Bharat Mission. Block Education Officers (BEOs) play a big role towards making this mission a success. We closely work with these BEOs to support them in developing and implementing innovative ideas for their blocks, based on student and school performance. BEOs surface challenges, discuss potential solutions among each other, and share their progress and achievements. Our role is to provide that platform, and support BEOs to lead relevant and impactful initiatives in their own blocks and communities. We learnt that we needed to:
4. Design programs that recognize the agency of people undergoing & leading change
5. Leverage existing people, processes, and policies so that existing systems are strengthened, and initiatives are more likely to sustain
6. Identify and give a platform to more champions, and design for celebration
We can see the impact at a small scale. How will we see impact at a larger scale? How do we attribute impact to our efforts?
This continues to be a tricky question which many of us grapple with across domains and geographies. We all recognize that as civil society actors, or catalysts for change, it is important for us to see that we are contributing to impact. Though this is important irrespective of the reach of initiatives, working at scale meant that we needed to find even more creative ways, potentially leveraging technology, to enable data visibility not just for interventionists, but also for the people leading and undergoing change within the system. When working at scale, especially in complex systems like education, we probably won’t be able to attribute impact seen in students/schools to our specific activities – and that’s something we had to learn to be okay with – it is the nature of complex and messy systems! In the Department of Education of a northern state for example, there were 700+ teacher mentors in the state who would regularly visit schools to support teachers. There were two key challenges however: mentors were not all on the same page with respect to what they needed to observe in schools and provide support on; and the second challenge was that data was either not recorded, or restricted to physical diaries of these mentors. Data and insights couldn’t come together to form patterns and trends which could help improve practices. We worked with the department in developing a supervision framework and tool for mentors, leveraging technology. What this achieved was that mentors not only now had a common understanding of their role, but they also had access to the data they were collecting – data which would inform their everyday action. This data would also then flow up all the way to the state level – where it could be analyzed without having to call or request or extract data from individuals. What we learn from this was how powerful it can be to:
7. Design for data transparency & enabling emission of data (where data automatically flows up to others, rather than a district officer, for example, having to individually check in and extract data from each mentor)
We can enable impact at a much deeper level if we work directly at a smaller scale
No debates there! What helped us was reflecting on how we’re looking at impact – if our definition of impact includes the end goal (student development) as well as a healthier system (education leadership), then it’s actually more impactful when we work towards enabling the existing system. While we may not see immediate and deep impact in working with the government at scale, we are able to catalyse system-wide incremental (+1) shifts which further have a ripple effect. When we were working with a smaller set of schools, we would be able to gauge the challenges and learning needs of teachers, and design teacher capacity building for them. At scale, however, the approach evolved to support existing institutions like SCERT (State Council of Educational Research and Training) and DIETs (District Institute of Educational Training). The role of these institutions already mandated teacher support – our role was in enabling better quality and frequency of this teacher support. In Bihar, for example, over 30 DIETs went on to develop relevant and need-based teacher capacity building modules for the teachers in their districts. The magical part is the ripple effect – now DIETs would be able to keep developing these modules whenever there is a need, and not depend on an external agency. We learnt that when we design for impact at scale, we need to:
8. Design for +1 shifts across the system
9. Design for ripple effects – enabling changes which have a multiplier effect
We don’t know the context of each school / unit of focus
…and neither can we ever know the context of every school / your unit of focus. One of the areas of improvement we focused on in our partnership with state education departments was improving community engagement in schools. Did we know the specific challenges of each school in engaging with the community? No. What we did know is that schools struggled to identify where they might engage communities and how. To address this, statewide campaigns on parent-teacher meetings were held, and school leaders were given guidelines or simple steps they could follow to design and implement effective PTMs. The steps were designed in a manner that enabled school leaders to choose strategies that worked for their schools. This was extremely powerful in enabling leaders to take decisions on what was best for their school, including – how parents were invited (street plays, handwritten notes), the timings of the PTM (some chose to conduct the PTM on a Sunday for better parent attendance, and some extended the PTM for 3 days). What was most helpful is that our role wasn’t to give these ideas, but rather to create a platform where these ideas could be exchanged – as there are always some bright spots and people who have found ways to deal with different challenges! We learnt that our role wasn’t just to propose initiatives and ideas, but also to:
10. Showcase bright spots so that they may be further spread to others
While it’s been a rocky ride, it’s been full of learning! Here’s a quick summary:
1. Design with impact at the center, not interventions
2. Think about what would work at scale, rather than scaling what works
3. Design for context aware solutions and recognize agency of those in the context
4. Leverage existing people, processes, and policies
5. Leverage partnerships
6. Identify and give a platform to more champions
7. Design for data transparency and data emission
8. Design for +1 improvements
9. Design for ripple effects
10. Showcase bright spots