Big data meets thick data
The tools that INGOs have traditionally used to make sense of the changing world are under strain. Donor reporting and legibility requirements increasingly call for quantitative indicators to demonstrate impact, which are ill-suited to the messy reality of many development challenges. And 2016 has shown that tools like polls, focus groups and surveys struggle to capture the sentiment and unarticulated needs of a population.
Initially hailed as a potential panacea to overcome existing limitations, big data implementation has proven to be challenging, if not, at times, harmful. Without proper handling and contextualisation, big data risks becoming deep fried data.
Paradoxically, then, the era of big data needs even more qualitative, granular knowledge of local contexts. A new approach is emerging, combining a growing interest in design thinking with the sector’s long-standing tradition of ethnography: the integration of “thin” big data with thick data. Development innovation labs are beginning to question what donors and governments see as “acceptable” evidence and are looking to bridge the quantitative vs qualitative divide.
In 2017, I predict INGOs will combine big data insights with user research and ethnography, thereby exploring the fuzzy “in between spaces” of data practice.
- We will see an increasing use of ethnography to understand how big and small data are used (or not) for decision-making and the rituals of resistance to data inside organisations. Studies like Reboot’s exemplary ethnography of data usage in USAID programming or IGC’s story of a government dashboard in India will increasingly become the norm. Combined with a growing awareness of behavioural science, this will finally dent the assumption “If only we give data to the right people at the right time, they will change their behaviour”.
- Interest in the social life of data will increase, with living labs, data collaboratories and “lives of data” explorations becoming more widespread. A focus on the distinction between upstream (reporting) vs downstream (learning) data will also generate insights on how to do data differently. Data infrastructures and the values they embed will be scrutinised.
- As the implications of data tracking become more obvious, INGOs will pay attention to practices of data obfuscation and how they affect their data collection and interpretation efforts.
- We will see more processes deliberately designed to merge design and data, as in Policy Lab’s data studio, Pulse Lab Jakarta’s haze field expeditions or the Urban Institute’s data walks to provide “real world insights” to big data analytics.
- There will be new attempts to harness digital tools to take ethnography to scale - whether to go beyond the limitations of surveys, provide context for policy-making, understand the emerging dynamics of collective intelligence or identify and understand what constitutes positive deviants and mapping local assets.
- Enabled by digital tools, donor reporting will increasingly see a hybrid of big and small data analytics with real time, qualitative stories from the field, where “Human stories are not the opposite of data”.