Smallholder farming systems contribute a substantial quantity of the food consumed in many lower and middle-income countries and contribute to the national and local economies. Despite the importance of smallholder farming, a transformation is needed in order to deliver food security and decent incomes for the farmers themselves and at the national level. This transformation must also be sustainable in terms of environmental impacts and social equity in order to be successful in the long term. The pressures of population growth, climate change, and land fragmentation compound the problem. Addressing these overlapping issues is a big challenge. One obstacle is the lack of good quality granular data linking these issues together. Household surveys are the workhorse method for gathering such data, but there are well-known problems that prevent household survey data from building up a “big picture” and delivering insights beyond the geographical boundary of each individual study. Such obstacles include the lack of access to datasets, differences in survey design, and respondent biases. Agile, data-oriented research tools can help to overcome these challenges. We use the term “agile” to imply methods that do not attempt exhaustive measurements, which are designed to be easy to use, and which entail some degree of flexibility in terms of adaptation to local conditions and integration with other tools or methods. Often these methods also nudge the behavior of tool users towards best practices. In recent years various research tools and approaches have been published which fit within our definition of “agile data-oriented research tools”. The domains these tools function in include monitoring and evaluation, intervention targeting, tailored information delivery, citizen science, credit scoring, and user feedback collection; all with the over-arching aim to improve data quality and access for those studying the sustainable development of smallholder farming systems. The goal of this Research Topic is to better define that niche, the ecosystem of tools and current practices, and to explore how such approaches can provide the underpinning knowledge required for the transformation of smallholder farming systems. One example of an agile data-oriented research tool is the Rural Household Multi-Indicator Survey (RHoMIS). It is a modular, digital system for building household surveys addressing the common topics in smallholder development. It was purposefully designed to give a broad overview of the farm system whist keeping survey duration to a minimum, to be user-friendly in implementation, and to be sufficiently flexible to function in a broad variety of locations and projects. Since 2015 it has been used by 30 organizations in 32 countries to interview over 34,000 households. The tool and database are open access and a community of practice is developing around the tool. We particularly welcome contributions that engage with the RHoMIS tool and data. However, we also describe the tool in order to provide an example of what is meant by an agile data-oriented research tool, and welcome contributions focusing on other tools or methodologies. We encourage the submission of manuscripts addressing the above topic, and those which fit within one of the following three sub-themes: (i) Perspectives or review articles which explore the niche, best practices, or promising approaches in agile data-oriented research tools for smallholder farm system transformation. Also, technology and code articles that describe new tools are welcomed. (ii) Original research articles presenting analyses based on data derived from agile data-oriented tools used at the project level. Examples include impact evaluations, adoption studies, targeting studies, or adaptive management, and should reflect on the additional benefit leveraged by the agile method applied. (iii) Original research articles that make use of the large amounts of data generated by such agile methods and/or link between agile data and other data sources. Examples include meta-analyses of data from multiple studies, layering data collected from different agile tools, or linking agile data to remote sensing or large-scale modeling outputs.
