A network connection is a design condition to investigate before choosing a platform. The connectivity gap raises questions about who can use a service, its cost, and what happens when a connection disappears. Global figures help frame those questions; local investigation should guide the answers.
A simple planning sheet can connect each essential action to its requirements: a device, power, a connection, an account, or another person's help. For each dependency, record a fallback to investigate and who will check it with users. Leave uncertain answers visible so that a prototype does not silently turn them into requirements.
Three-Quarters Online, 2.2 Billion Not
ITU's Facts and Figures 2025 estimates that about 6 billion people, roughly three-quarters of the world's population, use the Internet. Another 2.2 billion remain offline, compared with a revised estimate of 2.3 billion in 2024. The report places 96% of those offline in low- and middle-income countries. These are figures about people using the Internet, rather than a count of people living within reach of a particular network.
For a project team, begin with the intended task: reporting an observation, reading information, or checking a record. Ask how someone without an Internet connection would complete it. An account registration screen or a mandatory download can become the point where participation ends. Treat those steps as choices to justify during design, and include people without an existing connection in the discussion.
Where the Divides Run
ITU reports that 94% of people in high-income countries use the Internet, compared with 23% in low-income countries. The urban–rural comparison is also substantial: 85% of urban residents are online, against 58% of rural residents. Across the world, 77% of men and 71% of women use the Internet. Among people aged 15–24, use reaches 82%, compared with 72% of the rest of the population.
Each comparison describes a different grouping. Avoid treating a global percentage as a measurement of a particular village, campus, or household. Instead, use the differences to shape questions for community discussions: who has access to a device, who can use it when a task needs doing, and who would need another route? Ask whether the proposed interface suits the people expected to use it. A student team's own experience of connectivity should be an assumption to check.
Keep local access questions separate from the global statistics. Record whether a participant can use the proposed service when needed. Ask them to walk through the complete task on the device they would normally use, including any step that depends on someone else's phone or account.
Coverage Is Not the Same as Access
ITU estimates that 5G networks cover 55% of the world's population in 2025, including 84% in high-income countries and 4% in low-income countries. It also estimates around 3 billion 5G subscriptions, about one-third of mobile broadband subscriptions worldwide. Coverage, subscriptions, and individual Internet use describe different things. A subscription count should not be read as a count of unique users, and the presence of a network does not establish that a particular person can afford to use it.
Affordability deserves its own design questions. Despite a fall in the global median price of a data-only mobile broadband basket, ITU reports that it remains unaffordable in around 60% of low- and middle-income countries. The report identifies affordable services and digital skills alongside infrastructure as necessary for meaningful connectivity. When planning a service, ask about the cost of completing the whole task, including any repeated downloads or unsuccessful attempts. Check the assumptions behind an always-connected design before making it the only route.
Set a data budget for the task before adding optional media. Consider a text version of essential information and let users choose whether to download larger material. During testing, record what must be downloaded again after an interrupted attempt. Review those costs with participants alongside the time and effort the task requires.
Designing for the Offline Majority
Within a project intended to reach people who are mostly offline, define the essential task that should remain possible without Internet access. Consider saving an observation locally and sending it later. Show whether a record is saved on the device, waiting to send, or received by the service. Decide how to handle a retry so that a repeated action does not create confusing duplicate records. Make these states understandable to the person doing the work.
Consider SMS text messages or USSD menus as possible alternatives where the required mobile service is available; check their availability and cost locally. Neither is a substitute for establishing what users can actually access. Offer only the steps needed for the task, keep messages concise, and make any failure visible. For a shared device, consider how someone leaves an account and what information remains visible to the next person using it.
Test the design with a connection disabled, interrupted, and reconnected. Ask participants to explain what they believe has happened to their information at each stage. Include the person responsible for maintaining the system in that review, and document which functions require a network. A prototype should make its limits clear enough for users to decide what to do when a connection is unavailable.