Universal service fund planning · Zimbabwe first
Zimbabwe USF Compass
Where people still lack mobile internet, what would fix it, what it costs, and which projects the universal service fund should pay for first.
Zimbabwe in figures
People with no 4G signal
Modelled from towers, terrain and phone measurements.
Covered, but not online
A usage gap: affordability, devices and skills, not towers.
Connected with $5.0M of fund money
Ranked for most people.
Back-test hit rate
Ranking checked against later coverage.
Learn more
About USF Compass
Universal service funds hold money meant to connect the people the market leaves behind, but the evidence for where to spend it is scattered across operators, registers and reports. USF Compass brings it into one map and one ranked list, with the uncertainty shown rather than hidden.
Gap map
Every district and ward, coloured by why people are unconnected, with the towers, terrain and measurements behind it one click away.
Fund priorities
Pick a budget and a priority (most people, pro-poor, or schools and clinics) and see which projects the fund should pay for, and which operators can pay for themselves.
Every ward gets one of four labels
- Coverage gap (no 4G)No usable 4G where people live. Remedy: new sites, upgrades or satellite.
- Usage gap (covered, offline)4G reaches people but few use it. Remedy: devices, affordability, skills.
- Mostly connectedMost people are covered and online. The fund is better spent elsewhere.
- Unknown (not enough data)Too little evidence to say. Never counted as a gap by default.
The method
From gap to funded project
Four steps, each with its sources and assumptions open for inspection.
- 1
Find the gap
Where towers, terrain and phone measurements say 4G reaches, laid over where people live. Every ward is labelled a coverage gap, a usage gap, mostly connected, or unknown.
- 2
Choose the remedy
Upgrade an existing site to 4G, build a new site on a 3 km candidate grid, or bring satellite Wi-Fi to a school or clinic. Each comes with its backhaul and power.
- 3
Price it honestly
Capital, running cost and expected revenue as low, central and high ranges. The subsidy is the part operators cannot recover; zero means it is a case for licence obligations.
- 4
Rank and check
Projects ranked by people newly connected per dollar, weighted by your priority, then checked against history: would the ranking have found where 4G later arrived?
What we hold to
Our principles
Open data first
Public, citable sources drive the figures, so anyone can repeat the analysis.
Honest uncertainty
Ranges, not single numbers; areas without enough evidence are marked unknown, never as gaps.
Private by design
National registers stay behind sign-in. Public views show counts, never operator names or site codes.
Checked against history
The ranking is back-tested on what happened next, and the result is shown beside it.
Who it is for
Built for the people who decide and deliver
Fund administrators
A ranked, costed shortlist for a given budget, with the evidence behind each project.
Regulators
Which areas operators can serve on their own, as a basis for coverage obligations.
Operators
Where a subsidy closes the business case, and what backhaul and power a site would need.
Researchers and the public
Ward-level figures, open sources and a method written in plain language.
Open data
Our sources
Public datasets the figures are built from, each credited on the map.
- WorldPop population
- OpenCelliD towers
- Ookla open speed tests
- OpenStreetMap schools
- Copernicus DEM terrain
- geoBoundaries
- GridFinder power lines
- Meta Relative Wealth Index
- National internet usage rates
Questions
Frequently asked
What is a universal service fund?+
Most African countries collect a small levy from telecom operators into a fund meant to extend service where the market alone will not. USF Compass helps decide where that money goes.
Why only Zimbabwe for now?+
Doing one country deeply, with its own data checked, is worth more than a thin layer everywhere. The rest of Africa is shown with open data and is marked unknown where evidence is thin.
How accurate is the coverage map?+
The coverage model is scored against phone measurements it did not see, and the score is shown in the app. Where the model and measurements disagree, the area says so.
Can I see the national data?+
Counts are public. Points from national registers are shown only to signed-in users with permission, managed by the administrator.
See where the next dollar should go
Open Zimbabwe on the map and choose a budget.