Learnings from an Ecosystem Accelerator

This is the first of a series of articles where we share our learning from implementation of a system changes accelerator program. You can read more about the full program here. After each milestone, we will publish tips and tricks we used during implementation. We hope it will help other teams pursue systems change.

Although our research does not follow an academic framework, we place great emphasis on the quality of knowledge we produce, focusing on both language clarity and the value added to the field. As non-academics, our research experience may resonate with other non-profit staff members, aiming to create systemic change through programs like our accelerator.

First steps in recognizing the problem

The mapping of key stakeholders and the interviews is the state of the project aiming for system change, which allows us in-depth understanding of the problem. Key informant interviews combined with secondary data analysis, allows us to grasp the complexity of the issue, its root causes, and its symptoms. This stage of research also allows us to uncover hidden biases, which if unfolded could lead to a solution that doesn't include all the voices and perspectives.

We conducted approximately 100 interviews to better understand complex, interconnected systems of migration across Europe and Latin America. Ashoka’s extensive network of social entrepreneurs and innovators provided a strong foundation for this effort. Based on our network, previous research, and expertise, we built a diverse list of 500 key changemakers from both regions. This included individuals with lived migration experiences and those from second and third generations. We also engaged leaders from nonprofits and businesses, politicians, academics, media professionals, and content creators, ensuring diversity in gender, age, and perspectives.

The interview stage was a fascinating adventure, offering unique insights from key informants on the complex challenges of social and economic inclusion for people on the move. From this mapping exercise, we generated dozens of pages of insights, which have informed the accelerator process and resulted in two papers that will soon be available on our website.

Mapping: Building Your Knowledge Base

Mapping is the time to identify people you want to learn from. It is essential to start with a few contacts who can guide you to others, a process known as "snowball sampling." It is a very effective way to outreach to people that are outside of your bubble. However, it also brings risk of diversity distortion. During the mapping and interview phase you need to be very attentive and regularly check your list against diversity thresholds. Sometimes, as when you really create a ball out of snow, you need to make some corrections because it becomes too irregular - bigger on one side and smaller on the other. The same is in the research.

What is important is that in many cases you do not need to create an extensive demographic survey that becomes a boring part of an interview. When you talk to people and listen to them carefully you can easily spot those elements of their identity and social role that allows you to recognize if the overall group is diverse enough.

Discipline in Data Collection

A significant lesson from the mapping and interview phase is the importance of disciplined data collection. The data gathered will be analyzed, distilled into shared information, and ultimately transformed into knowledge. Since research often involves multiple team members with varied communication styles, interviews may differ in approach and focus. Align your team with core questions and goals to ensure consistency.

If some team members are less experienced in interviewing, consider pairing them with experienced colleagues for "job shadowing." Create spaces for team members to share and supervise each other’s work. Be mindful of emotional well-being, especially if the research touches on sensitive topics.

From Transcription to Insights

AI tools now provide efficient transcription solutions, but transcripts alone are insufficient—they are often lengthy and prone to errors. Interviewers should make detailed notes, summarizing key answers while maintaining the interviewee's voice and tone. Including direct quotes can preserve the essence of the conversation for analysts.

Internal validation sessions, where interviewers collaboratively summarize key insights, are invaluable. These sessions reveal recurring themes, patterns, and biases, enriching the overall understanding of the data.

Balancing Primary and Secondary Data

The mapping and interviews yielded valuable insights unique to our goals. However, we cross-checked this information with secondary data from academic institutions, NGOs, intergovernmental agencies, governments, books, and reports. While we didn't conduct extensive academic reviews, we relied on trusted sources and recommendations from interviewees.

Knowing when to stop is critical. Research should serve a clear purpose—in our case, equipping accelerator participants to develop innovative solutions. Revisit your research objectives regularly to avoid becoming overly absorbed in the topic. Especially if it is so fascinating and broad.

Simple Tools, Significant Impact

Our team used basic IT tools and AI to support the research process. Excel, transcription services in Teams and Zoom, and Miro for organizing notes were sufficient for our needs. While ChatGPT helped identify overlapping themes, the text was thoughtfully created by our team of experts. Maintaining high-quality outputs, even under tight timelines, is essential.

If you'd like to learn more about our llearnings from the system accelerator program, feel free to reach out to our team.