Photo: Baltistan Foundation’s field team

COVID-19: Why we need to listen to communities more than ever


*this post is shared in tandem with the UNDP of Pakistan

The ongoing COVID-19 crisis reveals that complex issues cannot be tackled by applying single point solutions. Unfortunately, despite the complex and interconnected challenges arising because of COVID-19, most of the current recovery initiatives are still designed as projects to address specific issues linearly.

Now more than ever, there is a need to create a mechanism to replace separate response processes with an integrated stream. After understanding the complexity of these challenges, the priority is to build a portfolio of options based on multiple collaborations for a systemic impact.

What is necessary is a deeper understanding of the social, economic and cultural dynamics that are conditioning the evolution of this crisis.

UNDP Pakistan’s Innovation-AccLab and the Agirre Lehendakaria Center (ALC) are collaborating to experiment a replacement for traditional linear projects, by co-designing a Social Innovation Platform to address socio-economic challenges — especially those which have emerged during the COVID-19 pandemic.

The Social Innovation Platform approach will allow us; to better listen to the communities, understand citizens’ perceptions and behavioural changes in real time, predict emerging changes in the COVID-19 pandemic, and contribute to co-design public policies.

The Social Innovation Platform fits well with UNDP’s socio-economic response to COVID-19, which focuses on avoiding “single-point solutions”

What does the Social Innovation Platform entail?

Social Innovation Platforms are integration engines that bring together a variety of actors, methods, and interconnected actions to allow for a more experimental and systemic approach in addressing the development challenges. Following are the core characteristics of the platform:

(1) new community listening tools and sense making processes,

(2) new co-creation and prototyping capabilities on five different levels (community actions, small and medium scale initiatives, large scale public/private partnerships, public service redesign, and new regulation),

(3) a portfolio approach to manage prototyping and scale,

(4) the possibility to set up locally focused investment funds.

Experiment: Social Innovation Platform in Gilgit Baltistan

Our first iteration of the platform approach experiment is being conducted in Gilgit Baltistan. Our partner, Baltistan Foundation, helped in deciding the geographical area for experimentation. We selected Hushe Valley — an area with unique geographical features, hard climatic conditions (only 2% of the land is cultivable), poor land connections and communication infrastructures. On the other hand, the area is located at the confluence of the Karakoram, Hindu Kush and Himalayan mountains, making it popular for its untapped tourism potential.

Photo: Jamil Akhtar

Pre-COVID19: How were we listening to the communities?

Active listening or deep listening is a pivotal skill that all ethnographic researchers and design thinkers use to comprehend people and their contexts. Building on Baltistan Foundation’s work that was on-going before COVID-19, AccLab Pakistan and ALC started mapping the system in different regions of Gilgit-Baltistan (Saling, Machulo, Hushe…).

We mapped out emerging themes, concerns, and interests. We then clustered these sets of insights into multiple categories of Agriculture, Livestock, Education, Skills, Health, Infrastructure, Tourism, and Employment. This provided us a frame to pin down the challenges and opportunities in the region as well as identify entry points for future interventions.

Pre-COVID ethnographic work in Gilgit Baltistan. Photos: Baltistan Foundation’s field team

Some initial opportunities identified pre-COVID-19 included; exporting opportunities of raw local products; livestock and poultry farming; organic agricultural production including local herbs (sea buckthorn); and overall tourism owing to the area’s unique landscape.

Ultimately, we are aiming at covering all existing complex narratives and segmenting their unified patterns, to visualize a portfolio of actions at different scales, that is connected to the communities’ needs.

Covid-19 outbreak: how did we pivot to digital listening?

In early March, in the midst of the listening work period, COVID-19 appeared. We had to immediately pivot our approach to continue our active listening process. We realised the importance of in-person information collection which should never be underestimated. However, because of COVID-19, we needed to find alternative approaches that would include adhering to safety precautions in our engagements while we continued our work. We collaborated with a local partner, VIAMO, to try and identify narratives, and responses to COVID-19, by using a wide variety of digital listening tools.

Examples of digital sources being experimented in Gilgit Baltistan

To maintain the element of ethnographic listening, we started experimenting with robotic calls from prominent local leaders to obtain information on peoples’ responses on COVID-19 in Gilgit Baltistan in real time. This increased our outreach, as we were reaching people in remote areas who are sometimes not included in the usual digital processes (e.g., people that can’t read).

How are we using Natural Language Processing for analysing the collected data?

All the information gathered from these calls is automatically transcribed into text and translated, to start the text-based data analytics phase (remember that we are still dealing with narratives). This is being done through Natural Language Processing techniques that provides us not only with real time tracking and narrative evolution patterns, but also sentiment analysis, keyword density, and bigram co-occurrences.

This digital analysis also helps cluster the narrative patterns through topic modelling, in the same manner, that we are segmenting the ethnographic information, thus allowing us to compare and integrate both.

Insights from NLP previous analysis in Thailand (Image courtesy of University of Deusto)

What does the near future look like?

As we will have the first results of digital data analytics soon, we have already started exploring ways to systematize this process and bring other elements of the platform to a digital space. Towards the end of the summer, we plan to digitally contrast this information with communities (i.e. Sensemaker) and start a co-creation process in September, based on the listening insights we have gathered and contrasted with local communities. We are working on a tool that visualizes, in an aggregated way, the mapping of COVID-19 socio- economic responses, key stakeholders, and networks.

By visualizing cultural element of the local context, we can create powerful tools that will allow us to better communicate such complex processes. This will also allow us to attract transformation capital on portfolio logics, as well as managing and evaluating the platform collaboratively, at an international level, which can be replicated in other Country social innovation platforms.

It is important to note that COVID-19 response initiatives and investments can lose their transformative potential if they are not interconnected as a portfolio of strategic investment options. This is why UNDP has designed a protocol to ensure that such initiatives are conceptualized and prototyped in line with this thinking. Innovating at the systemic level also requires tapping into new sources of financing as well as aligning resources in a different way. Social Innovation Platforms can help convene various sources of investment (public, private, philanthropy) and different investment tools (government programs, equity, social impact bonds, crowdfunding) that will provide capital into the entire innovation cycle iteratively (listening, co-creation, prototyping, scale).

By applying the social innovation platform approach, instead of a single point linear solution, UNDP Pakistan has become one of the few countries with more powerful and resilient tools for its programming to respond to a wicked emergency like COVID-19.

This blog has been co-written by Agirre Lehendakaria Center (ALC) and UNDP Innovation-AccLab Pakistan team. Together they are experimenting to develop Social Innovation Platform to tackle complex challenges, realize socio-economic transformation, and to help achieve the 2030 Sustainable Development Goals in Pakistan.



Regional Innovation Centre UNDP Asia-Pacific

Doing development differently through designing, developing, curating, collating and championing innovation and digital across the Asia Pacific Region.