How might we develop a pedagogical approach to social impact analysis that maps the content (successes & failures) and network connections across three core voices: the stakeholder, the beneficiary, and the government.
This case study discusses the background, research, problem, solution, and implementation of a pedagogical approach to social impact learning – AmplifyAI
2019 - 2021
Alan Hurt Jr.
Collective learning is the ability to build and pass on information over time. Given the rapid advancement of interactive technologies and projects aimed at capacity building, there remains a tremendous gap between how people and organizations collectively work towards solving social and environmental issues - i.e., how they collectively learn. A few of these issues include poverty & hunger, energy access, clean water, and farming.
The problems that afflict small-scale farmers and communities as a whole are often as chronic as the populations they serve. Because of this, it is important to be realistic about what pain points could be addressed within the constraints of this project.
Sitting on more than half of the world’s uncultivated arable land, it has been estimated that by the year 2050, Africa’s population will double. This poses quite a few problems for small-scale subsistence-based farmers and communities connected to these farms.
Traditional farming (conventional) or small-scale farming (individual farmers and cooperatives) is less likely to incorporate modern technologies to meet the growing food needs of Africa’s population. About 85% of farms in Ghana are under two hectares, and in Uganda 58% are smaller than one hectare.
With fewer adults farming, increased standards for food production, food produce, and a changing demographic, these communities will be pushed further into poverty. Understanding the need for a pedagogical Approach to Collective Learning — one that maps the content (successes & failures) and network connections through the lens of: small-scale farmers will be crucial in identifying areas for impact.
The challenge is that there is no scalable method to share or access the available solutions to small-scale farmers.
The main objective of exploratory research was to provide a starting point for acquiring knowledge into collective learning. Publications, media articles, original interviews, and other resources on a number of different topics in collective learning were analyzed.
On a macro level, this information was used to study how organizations learn. On a micro level, information was uncovered around the development of big data tools enabling collective intelligence learning and sharing of social impact knowledge in rural development.
The outcome of these tools is to bridge the knowledge divide between organizations (non-profit and non-governmental organizations) in the development sector. And thus, through our relationship to others, mediated by processes of intuition and transmission, we bring knowledge to life.
Before attempting to solve this collective learning problem, we attempted to validate our original hypothesis:
There is no scalable method to share or access the solutions available to small-scale farmers.
We framed the research around three goals:
Our focus was placed on uncovering qualitative data in the form of generative research in order to understand the competitive landscape. This analysis helped the team assess industry specific barriers.
From there, we conducted a competitive audit of existing software services designed for those working on collecting and sharing knowledge related to social impact. Many organizations have created technologies to solve one or a few problems associated with small-scale farming. No organization has presented a scalable method to share or access solutions that are available to small-scale farm holders as one complete solution. Following secondary research, we conducted one-on-one interviews with participants on the ground in Kenya and Tanzania.
Steppping out from behind the desk, the team recruited and selected a diverse sample of farmers to learn from. Purposive sampling allowed the team to cover the most dimensions in the group to achieve a diversity of opinions. 12 participants living in the Lake Victoria region of Kenya and Tanzania were selected and 6 were interviewed.
While the majority interviewed were farmers, we wanted to interview fisher-persons, general workers, and bicycle taxi drivers outside this focus for some comparison level.
Generative research allowed the team to uncover participants needs and frustrations around a key research question:
Is there a willingness by all parties to contribute to and nurture collective knowledge?
Once interviews were complete, the team began the painstaking process of synthesizing our findings.
We did this by organizing singles points of information into logical groups. Affinity diagramming is one of the best ways to bring order to what may otherwise appear to be chaos. This process allowed us to take disparate pieces of data and information to make the redundant essential.
After synthesizing our findings, three critical insights emerged, which helped to illuminate the points of tension within the farmer experience.
Deep needfinding is necessary to ensure that you are “doing the right things.” If this phase is bypassed, you run the risk of merely incrementally improving existing solutions instead of innovating.
And as my old granny once said, “Always be synthesizing.”
At the individual level, accumulating knowledge is difficult because learning is experimental.
We developed a fictional representation from the months-long research. Meet Alex.
“I see a lot of older farmers losing their lands to large companies and wealthier property owners. I need to know what technologies are out there so I can grow and sustain my farm.”
Alex is not simply a persona but a characterization of one of the types of people that we are designing for. This persona allowed the team to mobilize our future design around a knowledge space of thoughts and behaviors.
Once the research was completed, we attempted to understand our primary user by mapping observations collected from research.
Before examining any features or solutions, we wanted to check that all identified users' needs aligned with the business's goals. This was to avoid creating a solution for a pain point that was not in line with the company's focus. While the first user need mapped most strongly, all needs mapped to a business goal. At this stage, we decided to keep the top 3 in mind as we crafted our problem statement.
The goal of this process was to consider the experience of all the people affected by the product. This process codified our initial assumptions around why farming is so tricky.
A more efficient way to farm
Access to tools and methods to help me farm better
Access to and knowledge of certified seed, fertilizer, manure, etc.
To be incentivized to farm.
Collect and aggregate small-scale farmer data for further product development.
Create an MVP progressive web app for small scale farmers that enables collective learning and knowledge sharing.
Demonstrate with app usage, an increase in communities accessing and sharing collective learning knowledge and tools.
Collecting farming data (tools, best practices, and methodologies increases likelihood farmers grow and share knowledge.
Efficient farming begins with knowing what works and what doesn’t. A well connected farmer could lead to future financing.
More access to best practices and communities of learning may lead to increase in sharing tools and resources.
Based on the feedback from our business goals/user needs alignment, we prioritized our problem statements to ensure the design would focus on a primary problem.
The team was able to identify three core themes (and thus problems) to focus our brainstorming efforts towards.
Three core themes emerged:
Armed with our problem statement and papers full of sketched ideas, my teammate and I got to work, ideating some more, fleshing out the pieces that made sense, and figuring out how it would all fit together.
Instead of worrying about getting it right the first time, we can evolve a range of options, and rely on the structured process of prototyping and testing to test and improve our work.
How might we help Alex access information about farming practices/methods?
After the ideas were created, we took a critical look at each, validating or invalidating them based on what we knew about the project constraints, the business goals, and the user’s needs. The team arrived at a simple solution for maximum impact - tech stacks for social impact. The test Phase focused on Farming and communities of learning around farming methods, practices, and tools used. The solution can be scaled to become a database of working knowledge beyond farming - aimed at creating a running inventory of technologies created for the developing World - e.g. cookstoves, lanterns, and any other technologies addressing farming, human rights, water, HIV/AIDs, etc.
In considering how this feature would integrate with the rest of the app, several artifacts were created to explore and understand the app’s architecture, possible user flows, and entry points. A number of key inputs to conceptualization were considered - photos and ideas from the field and insights from our research synthesis.
Before making things visual, the app structure was mapped out. This helped to understand all screens that needed to be designed and what each screen’s relationship was to one another.
AmplifyAI can be used to inform decision making by mapping out the successes and failures andnetworks connections amongst stakeholders, beneficiaries, and governments.
The final solution replaces the thousands of websites, case studies, documents, and impact collaboration tools through the creation of a global database of data driven knowledge.
In our effort to understand knowledge, how Alex acquires knowledge, and what he does with the knowledge, the team identified a framework for representing knowledge - knowldege cards.
These cards divide knowledge into substructures. Unstructured and incoherent information (inputs) are connected and presented (outputs) in the form of cards (frames) by an information retrieval network (Alex).
In order to increase workflow and make it easier to scale and maintain the design, the team created a design system. The design system serves as a framework to map and represent information across the brand.
Careful attention was paid to how users discover information and how users access information.
An intuitive home screen provides Alex with all of the information he needs to make informed decisions about farming.
Multiple resource actors share knowledge as a basis for exchange. AmplifyAI will measure the activity's complexity and provide additional resources by developing knowledge techniques that preserve the elements being exchanged.
The hardest thing about this project was being the lead designer. I had to make difficult decisions around what to ship, what to compromise on and what we absolutely needed to get right.
Sometimes it felt like we were in a fast moving train but were also laying the tracks at the same time. AmplifyAI is a new product and the vision was to build a framework for disseminating knowledge . As a designer it meant I had to design for scalability and usability. At the same time, the interface had to be lightweight for the team to move at breakneck speeds especially during the phase of finding product market fit.
We ran into quite a number of usability and accessibility issues from testing that pushed more complex features to the backlog. I partnered closely with developers to address key issues and was in the room to make sure that technical limitations do not drive product decisions.
It was a challenge to get the engineering bandwidth to fully realize the design and feature set; however, it is fully functional. The final product provides a solid and usable experience, but a lot can be better in the UI, the performance, and representing knowledge.
Knowledge representation is a challenging field of study. The team leveraged the initial hypotheses to describe how to create systemic change by mapping a complex system to gain clarity.
Identifying specific points in the system where we can make a big impact is something I have done many times before, but developing a framework for learning and adapting over time as our system changes proved quite challenging.