Kenya is about to auction its fibre gap. Operators and their financiers should be in the room.
Kenya's ICT Authority has opened bidding for KDEAP's national backbone and cross-border fibre frameworks. The idea is simple: the government puts in only as much subsidy as each route needs, and the private sector builds, owns and runs the network. It could change how Africa pays for middle-mile connectivity.
I have read a lot of fibre tenders. Most of them follow the same script. The government designs a route, pays a contractor to build it, and then spends the next decade trying to maintain an asset it was never set up to run. The fibre gets laid. Too often, it doesn't get used.
The tender ICT Authority has just issued under the Kenya Digital Economy Acceleration Project (KDEAP) is different. It is one of the most commercially minded pieces of public infrastructure procurement I have seen on the continent. Anyone who builds, sells or finances fibre in East Africa should read it.
So what is actually on the table?
KDEAP is a World Bank-financed programme. Phase one is worth US$390 million, and its biggest component goes to digital infrastructure and access. This latest tender covers two parts of that component: pushing the national backbone down to ward level, and adding more fibre crossings to Kenya's neighbours.
ICTA is procuring it as two framework agreements. Lot 1 covers national backbone links. The sample list in the bidding document has around 180 routes and more than 11,000 kilometres. They run from Lodwar to Lokitaung, from Garsen to Mokowe, and along a chain of links across Mandera and Marsabit. There are also dense clusters in Kiambu, Nyeri, Kisii and Kakamega, plus metro builds in Nairobi, Mombasa and Nakuru.
Lot 2 covers cross-border links. There are 19 sample routes adding up to about 1,000 kilometres. They reach Tanzania at Namanga, Illasit, Holili and Lunga Lunga. They reach Uganda at Busia, Malaba, Lwakhakha and Suam. They reach Ethiopia at Moyale and Todonyang, and Somalia at Dadaab, Wajir Bor, Elwak and Mandera.
Every route uses at least 96-core cable. Lot 1 routes must be built within nine months of signing, and Lot 2 routes within six.
Why is the model the interesting part?
The routes are not awarded to whoever quotes the lowest build price. They are awarded through a multi-round reverse auction on an electronic platform. Pre-qualified operators bid for the smallest subsidy they need to make each route commercially viable.
ICTA expects that subsidy to fall somewhere between 1 and 50 percent of capital cost, depending on terrain, population and how commercially attractive the route is. The operator funds the rest, owns the network, runs it and carries the operating costs.
In return, the government gets dark fibre cores in proportion to its contribution. On cross-border links that is up to 42 cores if the full subsidy is claimed. Those cores become a permanent public asset for government, education and research traffic. They also act as a built-in guarantee that open access will be respected.
Think about what that does. Every shilling of public money is matched by private capital. Money goes further on the routes that are closest to viable, and is spent only where the market genuinely won't build on its own. The operator has every reason to keep the network running, because the network is its business. The subsidy is paid 20 percent at signing, 60 percent at partial acceptance and 20 percent at final acceptance. Service level agreements and a claw-back mechanism protect the taxpayer if things go wrong.
This is a first time for this blended finance model of public and private capital for digital infrastructure to be brought to market.
Who is this for?
It is not only for the giants. The entry requirements are demanding but reachable for any serious national player:
For Lot 1, a Tier 3 Network Facilities Provider licence. For Lot 2, Tier 2.
Average annual turnover of KES 200 million.
At least 300 kilometres of fibre built and operated.
A network operations centre with a track record.
A proper team of project managers, planners, engineers, technicians, and environmental and social specialists.
Joint ventures of up to three members are allowed.
The frameworks are also deliberately wide. ICTA can sign with up to 20 firms for Lot 1 and up to 8 for Lot 2, so there is space for more than the usual incumbents. Technical quality carries 80 percent of the evaluation weight and price 20 percent. That rewards operators who can show they will actually deliver.
One more point should interest readers of our last piece on power-grid fibre. The specifications tell bidders to prefer existing road reserves, pipelines, railway lines and electricity transmission corridors. Owners of that passive infrastructure can join a bid as sub-contractors, although not as joint-venture partners. So utilities, pipeline companies and other holders of rights of way are being invited into the market. The boundaries are starting to come down.
Why should banks and investors be paying attention?
Every route awarded under this framework needs private capital. The subsidy covers somewhere between 1 and 50 percent of the build cost. The winning operator has to find the rest, and it has to find it quickly, because the clock on a nine-month build starts at signing.
That is an opportunity for Kenya's commercial banks, development finance institutions, infrastructure funds and equipment vendors. It is a well-structured one, and here is why.
The public contribution is real money with a clear source. The subsidy comes from World Bank IDA credits and is paid directly under the Bank's disbursement rules. It arrives in fixed stages: 20 percent at signing, 60 percent at partial acceptance and 20 percent at final acceptance. For a lender, that is a timetable of payments from a highly creditworthy source, tied to engineering milestones that can be checked. It is a good base for bridge loans, construction facilities and receivables finance.
Much of the demand risk has already been handled. These routes were picked because they fill gaps in the national network: county and sub-county headquarters, border crossings and priority corridors. The auction sets the subsidy at the level that makes each route viable. Once lit, the fibre sits inside an open-access wholesale market with mobile operators, ISPs, enterprises and government as potential customers.
The borrowers will be credible. To be on the panel at all, an operator needs a Network Facilities Provider licence from the Communications Authority. It also needs audited accounts, KES 200 million in annual turnover, at least 300 kilometres of fibre already built and operated, and a working network operations centre. These are established businesses with assets and cash flow, not start-ups.
And the banks are needed from the very first step. Bidders have to prove they can access external finance, through bank guarantees, letters of credit or similar instruments. The threshold is at least KES 250 million for Lot 1 and KES 200 million for Lot 2. No operator gets onto the framework without a financial partner prepared to stand behind it, and the bank that issues that letter today is well placed to fund the build tomorrow.
The instruments needed are familiar ones:
Bid-stage guarantees and letters of credit.
Construction and bridge facilities, repaid from subsidy payments and later wholesale revenue.
Supplier and vendor finance for cable, ducts and splicing equipment.
Longer-term debt or equity for operators that win a series of routes and need to scale up.
Infrastructure funds looking for exposure to East African digital assets should also note that the framework runs for three years, with two more possible. An operator that performs well could win routes across several auction rounds and build a sizeable, long-lived fibre portfolio along the way.
There are risks, and a sensible lender will price them. Wayleave delays can push out build timelines. The claw-back and SLA provisions mean an operator that underperforms can lose subsidy. The 2028 project closing date is a hard stop.
But these risks sit with operators who have strong reasons to manage them, and the structure is transparent from the start. Next to most infrastructure lending on the continent, this looks unusually well defined.
My message to Kenya's banks is simple. Talk to your telecoms clients now, before the bid deadline, not after the auction. The operators who come to this with finance already lined up will bid more confidently, win more routes and build faster.
What has to happen to make this work?
Four things, and each is in reach.
The first is for operators to show up. A reverse auction is only as good as the competition in it. The more credible bidders on the panel, the further the subsidy stretches and the more routes get built. If you have been waiting for a well-structured way into underserved counties, this is it.
The second is that finance has to follow the framework. The subsidy only unlocks capital if capital is waiting to be unlocked. Banks, DFIs and vendors that line up with licensed operators before the auction starts will decide how many of these routes actually get built.
The third is delivery on time. KDEAP phase one closes on 31 October 2028. The build windows are short, and wayleaves will be the real test. County governments, KeNHA, KURA and the utilities will all need to move at the same speed as the contractors.
The fourth is that open access has to mean open access. The reserved government cores and the shared-infrastructure rules are the core of the public bargain. If smaller ISPs can buy wholesale capacity on these new routes on fair terms, the effect on last-mile connectivity will go far beyond the kilometres laid.
One more thing: who is this really for?
Kenya's Digital Master Plan sets a target of 100,000 kilometres of fibre. The gap between that target and today is not in Nairobi. It is in wards where a school has power but no broadband, where a county hospital has a server room but no backbone, and where a border town trades across the line every day but sends its data the long way round.
This tender is a serious, well-designed attempt to close that gap without asking the state to become a telecoms operator. It gives the private sector a fair price for going where it otherwise wouldn't, and gives Kenya's financiers a well-defined way to back them. That is exactly the deal the industry has been asking for.
Bids close at 10:00 EAT on 29 October 2026. They must be delivered in hard copy to the 12th Floor, Telposta Towers, Nairobi. Clarification requests are due 14 days before the deadline. The clock is running for operators and for the banks that will stand behind them.
Further reading
The bidding document (RFB KE-ICTA-538567-NC-RFB) can be downloaded free from icta.go.ke/tenders and tenders.go.ke. The auction rules are set out in the KDEAP Commercial Transactions Manual on kdeap.icta.go.ke.
About Digital Economy Advisors
Digital Economy Advisors is a Nairobi-based consultancy helping to accelerate Africa's digital growth. Our digital infrastructure practice works with operators, investors, utilities and governments on backbone, middle-mile and cross-border connectivity, from market and route analysis through to commercial models, blended finance structures and bid strategy. If you are a licensed operator preparing for the KDEAP framework, or a bank or investor looking to back one, we would be glad to talk. Call us on +254 733 400380, email info@digitaleconomy.ke
Comments