The data, formula and limits behind the Kenya Connectivity Map
I built the Kenya Connectivity Map to make the digital divide visible at county level. This post covers where the data comes from, how the score works and where it falls short.
The composite score
Each county gets a score from 0 to 100. Internet usage counts for 35%. Tower density counts for 25%, mobile speed for 25% and electricity access for 15%. I scale each layer to 0 to 100 before weighting. Internet usage carries the most weight because it is the most direct measure of access. Electricity carries less because it is a precondition for connectivity, and I did not want to count it twice.
The sources
Internet usage comes from KNBS and the 2022 Demographic and Health Survey. Tower density comes from OpenCelliD (March 2026). Mobile speed comes from Ookla Speedtest Intelligence (Q1 2026). Electricity access comes from the 2019 census. The grid and mini grid layers use the ENERGYDATA.INFO registry. School locations come from Giga, and school device counts from the ICT Authority DigiSchool dashboard. The full table is on the data page.
Where it falls short
Tower density is crowd sourced, so it can undercount towers in sparsely populated counties. Speed tests skew toward urban and higher income users, so rural speeds look better than they are. A few Starlink users can inflate a county average, which is why I flag Samburu and leave it out of the speed ranking.
The census electricity figure is from 2019 and the survey data from 2022. The map shows a snapshot, not a live feed. I update layers as new reports release.
Who uses it
NGOs, telcos, county governments and EdTech teams use county level data to decide where to put towers, mini grids and devices. If you need a county cut that the public map does not show, you can reach me through paulwamocha.work.
Each layer has its own write up: internet use, tower density, mobile speed and electricity access.
Frequently asked questions
How is the Kenya Connectivity Map composite score calculated?
Internet usage counts for 35%, tower density 25%, mobile speed 25% and electricity access 15%. Each layer is scaled to 0 to 100 before weighting.
Where does the map get its data?
KNBS and the 2022 Demographic and Health Survey, OpenCelliD, Ookla Speedtest Intelligence, the 2019 census, ENERGYDATA.INFO, Giga and the ICT Authority DigiSchool dashboard.
Explore the data
Open the interactive map or browse all 47 counties. County data is on the data page.