Picture a village that sits inside a mobile coverage map. The signal bars are full. And yet the clinic cannot send a referral, the shop cannot take a card payment, and a student cannot submit an assignment. Coverage, it turns out, is only one of several doors a person has to walk through to be meaningfully online. The ITU’s latest estimate — about six billion people using the internet in 2025, with roughly 2.2 billion still offline — is usually read as a simple split. The more useful reading is what sits behind it.
The number everyone quotes
ITU, the United Nations agency for information and communication technologies, publishes these figures each year from national reporting and modelling. On the surface the story looks like steady progress: the offline population keeps shrinking. But a single global percentage flattens very different situations. Being offline in a wealthy city because you choose not to use the internet is not the same as being offline in a rural district with no affordable data and no device.
Five gaps hiding behind one figure
1. Coverage. A network may exist on paper but not where people live. Rural and remote areas are the last to be built out, and the economics of serving sparse populations are difficult everywhere. When coverage data is measured by population reached rather than by where infrastructure actually performs, the gap is easy to miss.
2. Affordability. The UN Broadband Commission has long used a benchmark of entry-level broadband costing no more than 2 per cent of monthly income. In many low-income countries, a basic data bundle still costs a far larger share. Price, not signal, is often the real barrier.
3. Skills. Using the internet well requires more than a connection: reading, navigating interfaces, judging what is trustworthy. Digital-literacy programmes are chronically underfunded relative to network build-out, which is why some connected households still do not use the internet.
4. Devices. A shared phone in a household is not the same as a personal device. When several people depend on one handset, access is rationed by time, and women and girls are frequently the ones who wait.
5. Trust and safety. People avoid services they associate with fraud, harassment or surveillance. Online safety shapes whether a connection gets used at all.
Why a coverage map can mislead
Coverage maps are drawn from operator commitments and technical models. They answer the question “is there a signal here?” They do not answer “can this household afford to use it, does anyone know how, and do they trust it?” That is why countries can report high coverage and still report large offline populations. The two figures measure different things, and conflating them produces policies that build towers nobody uses.
If the barriers are layered, so are the remedies. Infrastructure still matters, but it works best alongside affordable pricing, public access points such as libraries and community centres, device programmes, and locally grounded digital-skills training. Regulation matters too: spectrum policy, competition rules and consumer protection all shape what a connection costs and how safe it feels. None of these is glamorous, and none works alone.
How to read the estimate
Treat the global figure as a headline, not a diagnosis. The interesting work is one level down: who is offline, where, and why. ITU’s own reporting separates coverage, affordability, skills, devices and security precisely because a single number cannot carry that weight. Anyone citing “2.2 billion offline” should be ready to say which of those five doors those people are standing behind.
Who counts as “online”?
Definitions matter more than they appear. ITU counts someone as an internet user if they have used the internet from any device, at any location, in the previous three months. That is a deliberately broad measure, and it means a person who checked a single message at a relative’s house is counted alongside a daily user. The estimate is a reasonable floor for “has used the internet,” not a measure of regular, productive access — a distinction that changes how much progress the headline actually implies.
The offline population is not evenly spread. It skews rural, older, poorer and, in many regions, female. Urban-rural gaps and gender gaps persist even where overall connectivity is rising, and they compound: a woman in a rural area with low income faces several disadvantages at once. Programmes that target only one dimension tend to leave the others untouched. This is why digital-inclusion strategies increasingly bundle connectivity with device access, training and safety, rather than treating network build-out as the single lever.
What to watch next
The trend is toward saturation in wealthy markets and slower progress where the barriers are layered. The useful indicator is not the total number online but the composition of the remaining offline group, and whether the gaps by income, gender and geography are narrowing or holding. A shrinking gap between connected and unconnected is progress; a shrinking gap that leaves the same groups behind is not.
This report draws on ITU’s connectivity statistics and the UN Broadband Commission’s affordability work.
Image: Syced · CC0 · via Wikimedia Commons.