Collecting Data on Social Enterprises 2025

Extrapolations and modelling While surveys can provide a flavour of the social enterprise community, they are not an instrument for establishing the scale of the social enterprise population in any given context. Where there has been a paucity of data, some studies have also utilized modelling or extrapolation methods to integrate the limited existing data into predictive models that can provide broader estimates. For instance, Siemens Stiftung published a study that drew on proxy data regarding the number of small and medium-sized enterprises and the rates of employment growth in various African countries to estimate prevalence rates and job creation potential among social enterprises.34 While such approaches necessitate close methodological scrutiny and can at best provide rough estimates, limitations on data availability mean that in some contexts modelling and extrapolation are the only possible options until new data collection initiatives are undertaken. Based on the combination of uncertainties in these calculations, it is important to note that these figures can only ever be viewed as rough estimates. However, they have often subsequently been used in communications, by politicians or advocates of social enterprise, and develop a life of their own – beyond the original researchers’ caveats. At times, these estimates and extrapolations have been scientifically questionable but have gained significant currency. Artificial intelligence and scraping While still a nascent approach, artificial intelligence (AI) and data scraping have also been used to extract and analyse data from the internet or other broad and potentially disparate sources. For example, Impact Intelligence utilized AI to analyse the story-form text of applications for a Bayer Foundation award on women’s empowerment in order to identify and analyse key challenges faced by female entrepreneurs in sub-Saharan Africa.35 Membership, networks, funders and financiers Membership and other support organizations often hold a significant amount of data, as many of these organizations have been operating in the sector for many years and have developed extensive networks. In addition, given the motivation of (potential) members to provide complete, accurate data in order to pass application screenings or to participate in various support programmes, they may have developed levels of trust and mutual understanding with members that enable them to achieve higher response rates with their surveys than academics or consultancies. Organizations collecting such data include the Aspen Network of Development Entrepreneurs, Ashoka, Catalyst Now, Echoing Green, IKEA Social Entrepreneurship, Impact Hub, the Schwab Foundation, Skoll Foundation and Yunus Social Business. However, data collected by some member networks often focuses on individual founders/ entrepreneurs and therefore is not quite as extensive with regard to organizational details. In addition, membership of some of the most high-profile global networks may be skewed towards particular types of entrepreneurs/ enterprises (i.e. high-growth oriented, more elite, technologically-savvy, highly-educated and well- connected) rather than being representative of the broader population of social enterprises. When comparing data across membership organizations, there is also a significant possibility of double- counting as certain individuals may be members or grantees of several organizations. Certification schemes Another source of data collected by membership organizations pertains to social enterprise certification – essentially the non-governmental equivalent of social enterprises’ legal status – and associated registries. These organizations collect and often periodically verify (the sometimes extensive) data to assess social enterprise status against various sets of criteria. Notable examples include the People and Planet First verification,36 which is facilitated through the Good Market digital commons. While various interlinkages exist between different certification and verification schemes (for instance, the People and Planet First verification is linked with Buy Social certification and Fair Trade Federation membership), they also vary widely in terms of the scope and specifics of the data they collect. Given that each certification is based on a set of core criteria specific to that particular verification scheme, the extent to which data alignment and comparability can be achieved through the basic certification surveys tends to be quite limited. Nevertheless, an important advantage of such schemes is that they often focus on specific practices instituted by social enterprises (often an area lacking in other surveys) and generally require applicants to provide concrete evidence to verify their claims, thus ensuring a greater degree of data accuracy than can be achieved through claims self- reported in surveys. Collecting Data on Social Enterprises: A Playbook for Practitioners 17