In today’s world, connection to the Internet is no longer a luxury: it is a fundamental necessity for participation in social and economic life. It is widely recognised that Information and Telecommunication Technologies (ICT) contribute to the achievement of the Sustainable Development Goals. However, connectivity alone is not enough: the true challenge lies in ensuring Universal and Meaningful Connectivity (UMC) — a vision that goes beyond mere access to deliver high-quality, affordable, and inclusive connectivity to everyone, regardless of their socio-economic background or location. This vision is at the heart of the UMC Project, led by the International Telecommunication Union (ITU), which seeks to bridge the digital divide and foster equitable digital inclusion. Achieving this vision requires a multidimensional approach to connectivity, addressing barriers such as affordability, quality of service, and digital literacy.

In the journey towards UMC, measurement plays a central role.  It is not possible to reach a goal without knowing where one currently stands. However, measuring the many dimensions of meaningful connectivity represents a major challenge. Effective strategies demand robust data, and herein lies a critical opportunity: the integration of innovative data sources alongside traditional methods to support the measurement towards UMC.

Traditionally, household surveys have been the cornerstone of data collection in measuring Internet access and use. These surveys offer invaluable insights, but they are often expensive to implement, especially in low-income countries.

Innovative (statistical) methods and data sources offer a transformative solution to these challenges, enabling policymakers to obtain more timely and granular data to inform their decisions. For instance, satellite imagery has emerged as a powerful tool for identifying underserved areas. By combining this data with demographic statistics, governments can pinpoint regions where connectivity infrastructure is lacking, plan for network upgrades, and even estimate mobile coverage across diverse terrains. This capability is particularly valuable in remote and rural areas, where traditional survey methods may face data collection difficulties.

Mobile phone data is another revolutionary data source. During the COVID-19 pandemic, governments of more than 40 countries partnered with mobile operators to analyse mobility patterns and monitor the impact of restrictions. Similarly, this data can be used to assess Internet usage patterns and validate the accuracy of mobile network coverage maps. However, accessing this data requires multi-stakeholder collaboration, a conducive legal framework that enables access to data while guaranteeing privacy and security concerns, as well as technical skills of the staff for acquiring, editing, processing and analysing the data, and the infrastructure to store, exchange and access vast amounts of data. National statistical offices and other stakeholders require capacity building to effectively adopt and apply these new methodologies.

Social media platforms and digital footprints also offer exciting possibilities for understanding connectivity dynamics. The way people interact online—what they search for, the languages they use, and the platforms they frequent—can provide indirect indicators of digital literacy and internet adoption. For instance, analysing social media trends (e.g. access to a blog, visualizations of media posted in social networks) can reveal how meaningful online content is for different demographic groups, shedding light on disparities in digital engagement.

The use of these innovative data sources is already demonstrating its value. Studies on meaningful connectivity in Brazil and Indonesia, supported by the ITU, demonstrated that estimates of Internet use obtained by processing mobile phone data were comparable to survey data.  analysing mobile phone data.

Internship programs at ITU for higher education institutions also offer opportunities to data scientists to work with a wealth of data. During an internship within the ICT Data and Analytics (IDA) Division of ITU’s Telecommunication Development Bureau (BDT), Damien Cornuejouls, a post-graduate student of Institut National des Sciences Appliquées de Toulouse, employed statistical techniques in particular small area estimation (SAE) methodologies, for estimating Internet access and use in countries with insufficient data. Using geospatial data from sources like ITU, Ookla, Meta, OpenCellID, and WorldPop, the project combined information on ICT infrastructure, internet speed, socio-economic indicators with detailed census data from Mexico to train linear models, random forest methods, and spatial models. The goal was to estimate internet usage in selected African countries — Ghana, Kenya, Rwanda — at detailed geographic levels. The research benefitted of previous experience in the use of Mobile Phone Data and big data sources by the Task Team of the UN Committee of Experts on Big Data and Data Science for Official Statistics chaired by the ITU, which has produced guidelines on how to use these sources for official statistics.

The study confirmed that data for Mexico can help estimate internet access in other countries, thanks to the high level of detail of its statistical surveys, and the heterogeneity of the small areas across the country in terms of socio-economic development, geographical features and ICT access and use indicators.  The results for Kenya can be observed by comparing Figure 1 (Proportion of Internet users from official data) and Figure 2 (Predicted proportion of Internet users estimated using a spatial model). For reference, the distribution of population density is shown in Figure 3.

ITU aims to expand the work using training data from a larger set of countries and with refined spatial modelling techniques. This research advances ITU’s mission to measure the progress towards Universal and Meaningful Connectivity by addressing data gaps through innovative statistical methodologies.

For more information on the project visit the website. For information about the use of innovative data sources, visit Big Data for Measuring the Information Society.