Decoding Delays: A Practical Guide to IATA Delay Codes

Johannes
Janning
Head of Partnerships
Jul 14, 2025
How a shared language, and real-time data from Cosmos, turns disruption into performance gains.
Why delay codes still matter in 2025
A flight is officially “on time” when it pushes back or blocks in within 15 minutes of schedule—the industry’s long-standing on-time-performance (OTP) benchmark. Every minute beyond that window chips away at passenger satisfaction and margins, adding up to an estimated US $67.5 billion in global economic loss each year.
To track root causes consistently across thousands of daily movements, the International Air Transport Association (IATA) maintains a standard library of two-digit delay reasons in Airport Handling Manual (AHM) 730.
A 1980s Standard Running Modern Ops
IATA’s AHM 730 delay-code table - first issued in the 1980s - remains the de-facto global language for on-time performance. Eighty-plus two-digit numeric codes (00-99) cover passenger, ramp, technical, weather, ATFM and “miscellaneous” causes. In 2011, IATA's AHM731 enhanced the AHM730 delay code setting by introducing sub-delay codes, allowing capturing operational nuance of operational burdens and pain points.
The delay reasons are processed as delay code within AHM780's Aircraft Movement Message (AMM) and AHM770's Aviation Information Data Exchange (AIDX), enables flight data exchange between airlines, airports, and third party service provider. Nevertheless, the variety of delay codes often cause friction in daily operations, as a correct assignment is sometimes not that obvious.
Why it still matters
Benchmarking: OEMs, airports and regulators reference the same taxonomy.
BLF & billing: Codes underpin station-level SLAs and penalty/bonus clauses.
Data pipelines: Most DCS/EDP systems still write 730 codes into flight logs by default.
Yet the schema shows its age in a world of EFBs, API-based data flows and collaborative decision making. No surprise that IATA is rolling out a new standard for delay code schema with AHM732, more on this later.
The ten families of IATA AHM 730 Delay Codes
Lead digit | Category | Typical examples | Collaboration hot-spots |
0 | Airline-internal (custom) | No gate available (06) | Turnaround planning |
1 | Passenger & baggage | Late check-in (11), baggage processing (18) | Check-in & baggage hall |
2 | Cargo & mail | Late warehouse prep (26) | Freight forwarder interface |
3 | Aircraft & ramp handling | Aircraft cleaning (35), fuelling (36) | Ramp-handler SLA |
4 | Technical & equipment | Aircraft change for tech reasons (46) | Maintenance control |
5 | Aircraft damage / IT failure | Damage during ground ops (52) | Safety reporting loop |
6 | Flight ops & crewing | Late crew boarding (63) | Crew rostering |
7 | Weather | De-icing (75) | Winter ops playbook |
8 | ATFM / authorities | ATC restriction en-route (81) | Network ops center |
9 | Reactionary & misc. | Aircraft rotation (93) | Disruption recovery |
(Full alphabetical list available on the Cosmos delay-code reference page.)
Where ground operations feel the pain
1. Passenger & baggage (Codes 11 – 19)
Typical culprits: late-connecting or paged passengers holding the door (15 PH), oversales seat reshuffles (14 PO), and bag-room backlogs when reconciliation slips (18 PB).
2. Aircraft & ramp handling (Codes 31 – 39)
Big three pain points: cleaning overruns (35 GC), fueling/defueling delays (36 GF), and loader or ULD positioning issues (33 GE / 38 GU). Each minute lost here eats directly into the scheduled ground time.
3. Technical & equipment (Codes 41 – 48 / 51 – 58)
High-impact events: unplanned defects (41 TD) or MEL-driven aircraft swaps (46 TC), plus ground-equipment damage (52 DG) and IT outages such as DCS failures (55 ED). A single 4x/5x event can unravel an entire hub wave.
Six best-practice habits for station and duty managers
In many stations, the biggest delay isn’t operational, but analytical. Poor or inconsistent coding means insights are lost. Cosmos helps improve code quality through structured feedback loops and change requests built into the workflow:
Tag every minute, even a five-minute slip, with the most precise code available. Granularity powers pattern-finding.
Validate in real time. Use handheld or cockpit EFB prompts to confirm codes while memories are fresh.
Separate primary vs. contributing causes. Cosmos lets you assign multiple codes per leg; the analytics know the hierarchy.
Close the feedback loop weekly. Sit down with your handler partner, review the top three recurring codes, agree on one action each.
Benchmark externally. Compare station performance to network averages; a “local normal” might still be an outlier globally.
Reduce miscoding by embedding AI delay code suggestions directly into workflows, whether via cockpit prompts, turnaround chat, or post-flight reconciliation.
Key takeaways
Delay codes are only as powerful as the transparency behind them.
When every stakeholder sees the same data in the same moment, corrective action trumps finger-pointing.
But AHM730 and 731 wont last forever. IATA's AHM732 is about to set a new benchmark for delay code schemas that airlines will have to adapt to in the coming years. It is designed to be simple and easy, and that's why we'll be taking a closer look at it in our next blog. Stay tuned.