Many travel agents in Africa and other emerging markets reach airline content through a GDS, and the major GDSs now carry NDC content from participating airlines. That leaves independent aggregators with a harder case to make, and leaves agents deciding whether a second connection is worth paying for.
The agent stays, the ticket is under pressure
The traditional travel agent has been written off for three decades and is still issuing tickets across most of the markets TDN covers. What has changed is the economics of the ticket. Airlines want to sell direct and through NDC, and they have used pricing to steer agents there. Business Travel News Europe reported that Turkish Airlines would charge $24 on EDIFACT bookings made through GDS channels from 1 October 2024, and named its own NDC aggregator partners after its split with Sabre. In Africa, Engineering News reported in 2024 that Airlink would move its trade transactions to NDC from 31 July, that trade partners would need to follow to receive richer content and preferential terms, and that Thomalex had been appointed as its first NDC aggregator partner. The message to an agent is similar in both cases: stay on the old rails and pay more, or learn the new ones.
The Turkish story did not end there. Travel Weekly reported that the airline returned to Sabre after a hiatus of nearly ten months, with Sabre displaying its NDC-enabled fares and ancillary products. An airline that leaves a GDS, builds its own channel and then comes back is not swapping one system for another. It is deciding which channel carries which content, and on what terms. In July 2026 Turkish Airlines also signed a new multi-year agreement with Travelport that includes NDC, extending a partnership that dates back to 1993. The NDC go-live is expected in a future phase, and Travelport said agencies would get a reduced surcharge for NDC.
Agents often compete on services that airline websites do not easily replicate. That means help with payment in places where cards are rare, a person to call when a flight is cancelled, and groups, visas and corporate accounts. A customer who pays in cash, by bank transfer or by mobile money, or who wants someone to chase a refund, is not an easy sale for an airline site or an AI assistant. That is analysis, not a measured figure. The more practical question is whether those agents can still get the content they need to sell.
The GDS now carries NDC
For many agents, the answer to that question is still their GDS. Amadeus reported 471.2 million bookings in 2024, up 4.7 per cent, which gives a sense of the scale of the system through which agencies reach airline inventory, and the large GDSs now distribute NDC content alongside EDIFACT. NDC remains a minority of that volume. When Amadeus reported its 2024 results, executive Decius Valmorbida told Business Travel News Europe that NDC’s share of its bookings would stay in the teens for quite a while. The Business Travel Magazine reported in August that Amadeus had renewed agreements with BCD Travel and American Express GBT that include NDC-sourced content, which shows where large buyers expect to get it.
Travelport’s July deal with Air Zimbabwe shows the same pattern from the airline side. The agreement returns Air Zimbabwe to global distribution for the first time in more than a decade and puts its content in front of Travelport-connected agencies worldwide. The announcement does not mention NDC, so it is evidence that GDS distribution remains commercially relevant to African airlines seeking international agency reach, not that NDC is flowing through it. For an agent, the practical point is that where an airline’s NDC content is integrated into their existing GDS, access may not require a separate aggregator connection, another booking interface or an additional technology contract. Commercial terms and surcharges can still differ.
Where the gap remains
In Africa the starting position is thinner than the GDS announcements suggest. A TPConnects and AFRAA survey published in 2025 found that 91 per cent of responses came from airlines that were not yet NDC-enabled. The survey recorded 70 responses from more than 25 airlines, so the exact number of airlines is not disclosed, and the sample leans towards smaller carriers, with close to 80 per cent of responses from airlines operating 15 aircraft or fewer. The finding describes airline readiness, not agents’ ability to reach NDC content through the systems they already use. It does show how much of the underlying airline market is still in transition, which limits what any GDS or aggregator can deliver.
Three gaps explain why a GDS connection does not settle the matter, and each depends on the airline rather than the GDS. The first is content parity. An airline can be available through a GDS without supplying all of its fares, bundles, ancillaries or promotions there, so coverage and content access are different things. The second is servicing. An agent has to change, cancel, refund and exchange a ticket after it is issued, and NDC implementations differ in what they allow. A supplier that gathers content but cannot service it reliably has not solved the agent’s problem, whether it is an aggregator or a GDS with thin NDC servicing. The third is cost. If an airline attaches a surcharge to bookings made through a GDS, as Turkish Airlines did, an agent has a reason to look at another channel.
What aggregators do next
Aggregators will not survive by repeating that NDC exists. The GDSs also aggregate, and they increasingly bring airline-direct and NDC content into their own distribution, so the real question is which model gives the seller and the airline more value. An aggregator’s position is strongest where it provides access to carriers or content that the major GDSs do not distribute comprehensively, particularly among African, MENA and low-cost airlines. Servicing is arguably its strongest remaining advantage, if it can make changes and refunds simpler for an agent than a GDS manages today.
Payments are the third area, and the argument needs care. Accepting mobile money is not a business on its own. African Pilot reported in June that Kora had joined IATA’s Financial Gateway, giving airlines and agencies using the platform access to African payment options including cards, bank transfers and mobile money through a single integration. That makes payment acceptance alone a weaker differentiator for an aggregator. The stronger proposition links payment acceptance to ticket issuance, settlement, reconciliation and support after booking.
The fourth is AI, where the evidence is thin. The plausible case is that AI agents favour suppliers whose content is structured and machine-readable, and an aggregator with clean connections to many carriers could become what an agent plugs into, unless the GDSs get there first. We have not seen test results we can verify, so this is a hypothesis. Smaller aggregators may not wait to find out. Airlines and agencies have commercial reasons to favour fewer, more capable connections. That creates pressure for consolidation, although how far the market moves in that direction remains uncertain.
What is left
The GDS carrying NDC narrows the aggregator’s case without ending it. What remains is the part a GDS cannot easily copy: the carriers it serves badly, the servicing it handles clumsily, and content the GDS does not carry comprehensively. An aggregator without those is selling access that an agent may already have.
For the agent the conclusion is less comfortable. NDC through the GDS will keep them able to sell, but it gives customers no reason to book through them. That reason has to come from what the agent does around the ticket, from payments to disruption to the trips they can build. The agents who last will treat the GDS as plumbing and compete on everything else.
