#867 2015 · Irembo (Rwanda Online) with the Government of Rwanda · e-government
Rwanda pays the company running its e-services only when a citizen finishes one
the problem
Government portals ship and then rot: the builder was paid at launch, so nobody profits from making them easier to use
background
Rwanda had spent years wiring government — national masterplans, a fibre backbone, rising connectivity. But digitising a service and delivering it are different problems. Around the world, standard procurement answers only the first: a contractor builds the portal, gets paid on delivery and moves on, while citizens who cannot find, complete or pay for a service quietly drift back to the paper queue. Portals built this way tend to stagnate the day they launch, because nobody's income depends on their working.
Rwanda's constraints sharpened the problem. Many citizens had no device or reliable connectivity, so a portal that only worked for the connected would serve a minority. The state wanted the platform maintained and improved for a decade or more, without an open-ended budget line. In July 2015 the Rwanda Development Board signed a 25-year contract with Rwanda Online, a private local tech company: design and operate a single-window portal for government services — the platform that became Irembo.
what everyone would do
The standard answer is to commission the portal on a fixed price or build it in-house, then police quality after the fact: audits, service-level KPIs, periodic rebuilds, a complaints desk. Every one of those levers arrives after the incentive has already been set. The builder is paid when the software ships, so the operator's income stops depending on citizens the moment the launch photo is taken, and oversight spends its energy chasing a misalignment it can no longer undo.
what they saw
The state's real product is not software but completed transactions. Paying per completion turns the operator's revenue into a live usability meter: every abandoned application is money the operator watches itself lose.
the move
The payment clause was the whole trick. Rwanda Online was not paid for building the portal — its income is a commission on every successful paid application, while the service fees themselves flow to the state. If a digitised service goes unused, the company earns nothing from it. So the operator keeps simplifying applications and maintaining services years after launch, and has built a nationwide network of thousands of agents who help citizens without devices or connectivity complete transactions — each completed one earning the company its cut.
why it works
Because revenue arrives only when a paid application is completed, the operator's income becomes a function of citizens' success rather than of code delivered. That turns every confusing form, unnecessary document upload or broken payment step into a direct hit on the company's own earnings, so simplifying services and maintaining them years after launch — the period when government-built portals usually stagnate — is permanently in the operator's financial interest. The state did not need to supervise usability; it priced usability into the contract and let the operator's own ledger do the supervising.
the payoff
Government figures for 2025: 240 services, 120M hours saved, 80% faster. The operator's own site reports 106 services, 40M+ hours saved.
where it breaks
The model breaks where completion is not the operator's to control: if ministries delay approvals on the back end, the operator starves on revenue it cannot earn, and if the state can redirect fees or revoke the concession at will, the long-horizon incentive collapses into short-term rent-seeking. It also funds itself only on services citizens actually pay for — free services, or usage too thin to cover costs, need another revenue source. And where usage is low for reasons no redesign can fix, the same mechanism that punishes friction simply starves the operator: that is the point of the design, and its sharpest risk, at once.
what came after
Rwanda Online later rebranded as Irembo; the government holds a stake in the company, and in February 2026 IremboGov was named Best Government Service in the World at the GovTech Prize during the World Governments Summit in Dubai. The completion-commission clause remains Rwanda's answer to the question every e-government project eventually faces: why does the portal keep improving years after launch?
references
- [1]Untold Story of Irembo, Rwanda's One Stop Centre for Gov't ServicesKT Press, 2019ktpress.rw
- [2]The Irembo model in RwandaPublic Digital, 2022public.digital
- [3]Irembo — company site, current impact figuresIrembo, 2026irembo.com
- [4]Irembo Platform Transforms Access to Public ServicesRwanda Dispatch, 2025rwandadispatch.com
- [5]Rwanda Wins Best Government Service in the World Award in DubaiGovernment of Rwanda, 2026rwandainuae.gov.rw