How Multi-Source Travel Content Works
How multi-source flight search works across GDS, NDC, low-cost and consolidator channels: aggregation, normalisation, deduplication and live or cached fares.
How multi-source flight search works across GDS, NDC, low-cost and consolidator channels: aggregation, normalisation, deduplication and live or cached fares.
Multi-source travel content means one search that queries several separate supply channels at once: GDS fares, airline NDC offers, low-cost carrier connections and sometimes consolidator net fares. Calling every source is the easy part. The real work comes after: normalising each source's format so results can be compared, deduplicating the same flight offered through two channels, and tracking which sources return live availability and which return cached data. More sources help a travel agency only when all three steps are handled well.
Multi-source travel content is one search box backed by several separate supply channels. When a traveller or an agent searches for a flight, the platform behind the box may query several sources at once. Those can include a GDS, airline NDC connections, low-cost carrier connections and sometimes a consolidator. Sending those queries is the easy part. The quality of the results depends on three later steps: making each source's data comparable, recognising the same flight offered twice, and knowing which answers are live and which are cached.
Definitions.
Flight content is split because distribution grew as a set of separate commercial relationships, not one shared database. Full-service airlines have long distributed through GDS agreements. Many now also sell through their own NDC APIs, and each airline builds its NDC offer in its own way. Low-cost carriers often sell mainly through their own channels and direct connections. Consolidators add negotiated net fares on top, ticketed on their own accreditation.
None of these channels was designed to work with the others. A seller who wants all of them in one search box has to bridge the gaps. The comparison of GDS, NDC aggregators, consolidators and flight APIs explains who settles and services the ticket in each model. For the GDS and NDC split on its own, see NDC vs GDS.
A typical multi-source flight search draws on four kinds of source. Each one answers a search differently, refreshes at a different pace and has its own idea of what "available" means.
| Source | What it usually adds | Data shape | Availability | Commercial model |
|---|---|---|---|---|
| GDS | Broad full-service airline coverage | Standardised fares and fare rules | Live or cached, depending on setup | Published and private fares, commission or fees |
| Airline NDC | Bundles, extras and airline-specific offers | Offers and orders, different per airline | Usually live from the airline | Airline's own pricing and conditions |
| Low-cost carrier connection | Low-cost fares and extras | Each carrier's own format | Usually live | Carrier's own fares, often paid by card |
| Consolidator | Negotiated net fares | Consolidator's own fare files | Filed fares, with seats checked live or from cache | Net fare plus the seller's markup |
The table is a simplification. Some low-cost carriers also sell through a GDS, and some GDS setups use cached shopping data. Even so, the pattern holds: four sources, four formats, four sets of rules.
Aggregation sends one search to every relevant source and gathers the answers in one place. In practice that means calling each source in parallel, setting a time limit for slow responses and merging the lists that come back. The step is well understood, and most platforms do it.
Aggregation stops at "collect". Merging the lists only works if every list means the same thing, and at this stage they do not. One source quotes a price with bags included, another without. One labels a cabin "premium economy", another uses a booking class code. Showing these side by side is possible. Sorting them fairly is not possible until the next step.
Speed is the other limit. A search is only as fast as the slowest source it waits for. Platforms usually set a cut-off and return what has arrived, which means a slow source can quietly drop out of results.
Normalisation converts each source's fare rules, baggage allowances, currency, cabin labels and fare basis codes into one internal structure. A GDS fare, an airline's NDC offer and a consolidator's net fare can describe equivalent products in three structurally different ways. Normalisation is what lets a platform compare them.
The work covers several layers:
Until results are normalised, "cheapest first" means little. A fare that looks cheaper may simply exclude a bag the other includes. Normalisation is unglamorous work, and it decides whether results can be trusted.
The same flight appears twice because the same seat can be sold through two channels at once. A GDS may carry the published fare, while a consolidator offers a net fare for the same flight. An airline's NDC offer may bundle a bag that its GDS fare leaves out. Each version has its own price, conditions and settlement path.
Good deduplication recognises the overlap and decides what to show. Usually the platform keeps the offer that gives better value under the fare rules that apply to it. Sometimes both versions should stay, because a branded NDC bundle offers something the plain fare does not. Poor deduplication does one of two things: it shows obvious duplicates, or it silently drops a valid option.
Deduplication is also where source quality shows. A source that keeps producing near-duplicates with inconsistent fare rules adds noise to results without adding real choice.
Live availability comes from querying the source at the moment of search. Cached availability comes from a stored copy refreshed on an interval. Cached data is faster and cheaper to serve. The trade-off is accuracy: a fare that looked available in results can be gone when the customer tries to book.
Combining live and cached sources makes the problem harder, because the results look identical on screen. A multi-source platform needs to track internally which results are live and which are cached. Before booking, it should recheck price and availability with the source. If it does not, the failure shows up later as a booking that does not confirm or a price that changes at checkout.
For agencies, the symptoms are familiar: a customer complaint about a vanished fare, a booking stuck in pending, or a price difference the agency has to absorb.
A travel agency should judge a multi-source platform on how it handles content, not on how many sources it lists. Five questions cover most of it:
The answers matter more than the logo slide. A platform that can explain each step clearly is usually one that handles it.
For a travel agency or OTA, the source count is the least useful number to ask about. A platform with fewer sources, each handled well, often gives better results than one with many sources merged loosely.
Multi-source content also changes the work after booking. Changes, cancellations and refunds follow each source's own process, so ask how servicing works across sources too.
altovo (formerly easyGDS) is a travel commerce platform. For travel agencies, altovo puts GDS, NDC and low-cost content behind one search, with 650+ airlines bookable, including NDC (as of Oct 2026). Across flights, hotels and more, the platform pulls from 1,000+ content sources and returns one clean set of results. The platform is Level 4 NDC certified and sits alongside existing GDS contracts, so an agency can add NDC and low-cost content without replacing what it already has. Agencies can add their own contracted supply too.
The story of how altovo built unified inventory describes the merge. When two suppliers return the same rate, one is kept. When NDC returns a branded bundle a GDS cannot represent, the bundle is kept. altovo's median processing time on that merge is 120ms, and connectivity runs with a 99.99% uptime SLA. See altovo for travel agents, the inventory and connectivity pages. Self-serve travel agent accounts can be live in under 10 days, from sign-up to live bookings.
Want to see what sits behind one search for your agency? Leave your details and our travel agent team will set up a 30-minute walkthrough.
Book a 30-minute walkthroughConnecting is possible, but it is only the start. GDS, airline NDC APIs, low-cost carrier connections and consolidators each use their own data format, fare rule structure and update cadence. Each connection also needs upkeep as suppliers change their APIs and content. The larger ongoing work is keeping every output comparable in real time, so a result means the same thing whatever its source.
Normalisation converts each source's fare rules, baggage allowances, currency, cabin labels and fare basis information into one consistent internal structure. Without it, a platform can show results from several sources side by side, but it cannot rank or compare them honestly, because prices can only be compared when every result states what it includes in the same terms.
The same seat can be sold through two commercial channels at once, for example a published GDS fare and a consolidator's net fare for the same flight. Each has its own price, rules and settlement path. Deduplication logic has to recognise the overlap and decide what to show. Usually one offer is kept, but both can stay when each gives the customer something different, such as a bundle with a bag.
Only when the new source is normalised and deduplicated against the sources already connected. A source that is handled poorly adds latency, duplicates and stale fares without adding anything a customer can book. Source count and content quality are different measures, and agencies should ask about both before choosing a platform.
As fresh as each source allows. Some connections return live availability on every query, while others serve cached results refreshed on an interval. A multi-source platform needs to know which is which and recheck price and availability before booking, so a customer is not quoted a fare that has already gone.