Akvo: The next generation of data collection for sustainable development

Cisco is proud to support Akvo, which provides data services and tech solutions to development partners in the water, agriculture, and climate sectors. We know that data is critical for effective decision-making, collaboration, and accountability. Akvo supports agencies and governments around the world to improve their programs so they can achieve lasting impact.

This post comes from Joy Ghosh, Akvo’s Tech Lead. Joy has been working in the field of data and technology for development for the past 10 years, and specializes in building data platforms for the agriculture and water sectors.

---

Here’s how data is usually collected in the development sector. Let’s say a project is set up to improve the income of smallholder farmers or achieve safe water access in a certain region. Enumerators are sent out to collect data on the communities which are impacted by these issues. Interventions are then put in place to improve the outcomes and contribute towards impact. At the end of the program, enumerators are sent out once again to collect data and monitor change. But what if we could supplement this tried and tested method of data collection with continuous input from the field, rather than just baseline data at the beginning and the end of a program? And what if, rather than knocking on doors to collect data, community members had a platform to share information on their challenges and needs. What if we could use data collection as a tool for continuous community engagement?

With support from the Cisco Foundation, we’ve been developing technologies to make this a reality. Unstructured Supplementary Service Data (USSD) and phone-based data collection is often seen as a cost-saving alternative to field data collection — ideal for situations in which you can’t go into the field due to health or security emergencies. But this technology should be seen as a complement to in-person field data collection, and the benefits go far beyond reduced costs and security challenges.

A hybrid approach to monitoring

Today, enumerators are sent out into the field to find respondents from whom to gather data. This traditional method of data collection has clear benefits – enumerators can observe first-hand what’s going on in the field, and project managers receive comprehensive, quality datasets. However, there are also numerous drawbacks.

In-person data collection is resource-intensive and therefore periodical. The lack of continuous, timely data means that often, interventions are detached from the realities on the ground. Field-based data collection provides snapshots into the current situation, rather than continuous feedback that allows program managers to fill in the context, and adjust strategies based on evidence. Using a hybrid approach to data collection – remote and in person – could help to complete the picture. Program managers can use remote data collection to gain a broad understanding of a specific problem, and then in-person data collection to focus on potential problem areas. Alternatively, in-field data collection can be conducted at, for example, annual intervals, and then remote data collection can fill in the gaps and provide context continually. This would allow program managers to see change as it occurs, allowing for more timely responses, targeted interventions, and better results.

In-person data collection is also a physically exhausting process. The sheer logistical challenges may lead to enumerators collecting data from 10 people in one village and none from the surrounding five, resulting in a skewed sample. With remote data collection, you have the potential to exceed your sample, reaching a much wider geographical scope with significantly less resources.

Finally, in-person data collection makes it difficult to seek out local change makers in the community, and brings to the fore questions of data ownership. What are the incentives for respondents to give away their time and their data? Does the data still belong to them? And how, if at all, does the data get fed back into the community? Innovations in remote data collection technology could allow for a more citizen-centered approach.

Harnessing the power of remote in Kenya

In Kenya, Akvo created a system for distributing, accessing and responding to surveys via radio and text messaging in cooperation with county governments of Kajiado, Laikipia and Kwale, Central FM Kenya Radio, and Africa’s Talking, a local USSD aggregator. Akvo and Africa’s Talking jointly produced an Application Programming Interface (API) integration that allows surveys, traditionally administered by a mobile surveying application, to be received and completed via text message. We partnered with Central FM Kenya Radio to establish a schedule for broadcasting USSD codes to listeners, analyzing and interpreting data alongside Health Authority Officials, and distributing survey results.

Radio DJs that host interviews with representatives of Kenya Water for Health Organization received data process training and then broadcast the USSD codes to listeners. Each time a survey was announced, the DJ would explain that answers from completed surveys would be automatically sent to county databases and live data visualizations (maps, charts, and graphs) would be updated. DJs actively distributed survey codes, observed responses coming-in, updated live data visualization dashboards, interpreted results in conjunction with Health Officials, and communicated updated information back to the community regarding two thematic areas: integrated water management and gender-based violence.

This approach allows organizations to ask the community directly for input, be it in relation to their water facilities, healthcare, food security, or any other societal issue. This has a host of benefits, including:

• Raising awareness around the issue at hand by broadcasting it
• Gathering data at regular intervals via broadcast for a fraction of the cost
• Identifying individuals in the community who care about and actively participate in community issues, because they come to you, rather than you going to them
• Introducing incentives for feeding in data continuously and on a timely basis (through gamification, for example)
• Feeding data directly back to the community — they are aware and involved in the problem and empowered with opportunities to contribute to the solution

This pilot is one of many that we’ve been testing in various sectors and contexts with support from the Cisco Foundation, and shows how traditional data collection methods in the development sector can be innovated to improve results. At the time of writing this blog we are in a scaling phase, aiming to support an additional 10 partners (government and organizations) this year with crowdsourcing solutions such as USSD.

What’s next for the sector

Just like the switch from pen and paper data collection to smartphone-based data collection, this evolution will take time and learning. There are numerous benefits to remote data collection if used in the right way and in the right context, but we have to learn which methods are best suited to which contexts, and what the drawbacks and challenges are. Beyond that, we have to think differently about how data should be collected based on the different mediums. For example, you can’t run the same survey on USSD as you would in an in-person survey because it would be too long and too cumbersome for the respondent. This is where design thinking comes into play, and program managers consider what data is needed, how  that data is best collected, and what obstacles there may be.

We’re continuing this work to identify the best practices, challenges, and opportunities, with financial support and technical advice support from the Cisco Foundation.