Across the 11 markets where we operate, the publishers running on our infrastructure deliver more than ten billion impressions a month. The vast majority of those impressions clear at fill rates that would make a US ad-ops director resign in protest. The reason is not that nobody wants to advertise here. The reason is that the plumbing between buyer and seller has been missing for two decades. Here's what we've learned trying to build it.
The headline gap
The IAB's most recent benchmarks put US digital audio fill at around 78% across the major streaming and podcast networks. Our continental average, weighted by impressions, sits closer to 34%. That's not a rounding error. That's a 40-point delta on roughly the same per-listener engagement, the same dwell times, and in many cases the same global brands buying the inventory.
For most of the last decade, the standard explanation has been some combination of three things: African audiences are too fragmented to measure, ad budgets are smaller in absolute terms, and the local market doesn't have the infrastructure to support programmatic buying at scale. All three are partly true. None of them is the actual problem.
What the gap actually is
The actual problem is that the buying layer has been built for transactions that don't match how money moves on the continent. The typical DSP expects a buyer with a US dollar credit card, a fixed-IP campaign manager, monthly billing in dollars or euros, and ad standards designed for a single-language broadcast environment. Almost none of our publishers' real demand looks like that.
A typical month for a Lagos radio group looks like this: 60% of demand comes from local agencies billing in naira, 25% from continental brands billing in dollars but paying through subsidiary entities in each market, and 15% from global brands buying through agencies of record in London, Johannesburg, or Dubai. Each of these flows has different settlement timing, different creative spec requirements, and different reporting expectations. The standard ad-tech stack handles exactly one of them well.
The three things that have to be true
From the publisher side, three operational facts have to hold before fill rates start to converge with global benchmarks:
- Inventory has to be sellable in the buyer's currency and on the buyer's payment timing. If a local agency can only pay in naira on 60-day terms, a buying layer that requires dollar-denominated NET-30 is structurally not going to clear that demand.
- Creative trafficking has to handle local language and dialect-level adaptation without the publisher doing it manually. A campaign in Lagos that needs to run in Yoruba, Igbo, and Pidgin variants cannot require the publisher to traffic each version as a separate IO.
- Reporting has to be reconcilable against the buyer's internal expectations. Most publishers can produce a delivery report. Very few can produce one that reconciles cleanly against an agency's internal media plan in the format the agency's compliance team needs.
Each of these sounds operational rather than strategic. That's because it is. The fill-rate gap is not a measurement problem and it is not an inventory problem. It is a back-office problem dressed up as a market problem.
"Africa is not under-monetized because the audiences aren't there. It's under-monetized because nobody built the boring infrastructure to clear the trades. That's the work." Mac Maison, founder & CEO
What closes the gap
The work we've spent the last six years doing, in the simplest possible terms, is building one operating layer that can absorb the way money actually moves in African ad markets and present it to publishers as a single, clean fill engine, and to advertisers as a single, clean buying interface.
On the publisher side, that means handling multi-currency settlement, multi-language creative trafficking, and reconciliation reports in the format each demand source needs, all behind a single rev-share contract. The publisher gets one statement at the end of the month showing inventory delivered and revenue earned, denominated in their local currency.
On the advertiser side, it means a buying surface where a brand can plan a campaign across 11 markets in 40+ languages, with consistent fraud controls, consistent measurement, and one IO covering the whole network. The buyer doesn't need to learn how naira-denominated billing works in Nigeria, or how matatu screens are licensed in Kenya, or which podcast network in Accra handles dynamic ad insertion. They just buy the audience.
The early evidence
Across the publishers who've been on our infrastructure for 18 months or longer, average fill rates have moved from roughly 18% at onboarding to 67% as of last quarter. That's not the US benchmark yet. But it closes more than half of the gap, and it does it without changing anything about the underlying inventory or the audiences listening to it. The audiences were always there. The infrastructure wasn't.
We have a lot more to learn. The next markets we're opening — Ethiopia, Mozambique, and Côte d'Ivoire among them — will each have their own version of the back-office problem, and our system will have to absorb them too. That's the work. It's not glamorous, and it doesn't show up on a panel deck. But it's what actually moves the number.