When an airline announces a shiny new route, the press release talks about growing demand and exciting destinations. What it never mentions is the two years of argument that preceded it: the demand models, the slot negotiations, the treaty checks, the fleet Tetris, and the internal fights between planners who believe in the city and finance people who do not. A single daily long-haul flight can commit an airline to well over a hundred million dollars of aircraft time, fuel and crew over a few years. Route planning is where airlines place their biggest bets, and it is one of the least understood corners of the business.
Step one: Measuring demand you cannot see
The foundation of every route decision is a deceptively hard question: how many people want to travel between city A and city B, at what price, and how many are already doing it with a connection? Planners start with booking data from reservation systems and government traffic statistics, then estimate the true origin-and-destination demand hidden inside connecting itineraries. A market like Columbus to Amsterdam may show almost no nonstop passengers, because there is no nonstop, yet thousands of people make the trip every year through other hubs. Finding those buried flows is the planner's treasure hunt.
The workhorse tool is the QSI, the Quality of Service Index, a model that predicts what share of a market each airline will capture based on schedule quality: nonstop beats connection, morning departure beats red-eye, big fast jet beats small slow one. Feed a proposed schedule into the model, and it forecasts passengers and revenue. Academic groups such as the MIT Global Airline Industry Program have published extensively on these forecasting methods, and every network airline runs a proprietary version. The models are good, and they are also routinely wrong, which is why the industry graveyard is full of routes that looked brilliant in the spreadsheet.
Every route map is a set of bets, and the spreadsheet never gets the last word; the passengers do.
Step two: Matching the airplane to the market
Demand means nothing without the right tool. Range and seat count define what is possible: a 180-seat A320neo cannot cross the Pacific, and a 350-seat 777 is ruinous on a route that fills 120 seats. The revolution of the past decade is the long-range twinjet with modest capacity, above all the Boeing 787, which was explicitly designed to make "long, thin" routes viable, city pairs with real but modest demand that could never fill a jumbo. Boeing's own Commercial Market Outlook credits this class of aircraft with opening hundreds of new nonstop city pairs that previously required a connection. Perth to London, Auckland to New York, and the ultra-long-haul experiments of Qantas's Project Sunrise all exist because the machines finally allowed it.
Cargo quietly shapes these choices too. A widebody's belly holds freight that can contribute a meaningful slice of a route's revenue, so a city with strong export flows, think pharmaceuticals or electronics, can tip a marginal passenger route into profit. Our feature on how cargo airlines move the world explains why planners always check the freight forecast before the passenger one.
Step three: Permission, slots and politics
International flying is not free. It runs on bilateral air service agreements negotiated between governments under the framework of the Chicago Convention, administered by ICAO. Some treaties are liberal open-skies pacts; others cap frequencies, name specific airlines, or restrict routes, which is why some obvious markets go unserved for decades. Then come slots. At coordinated airports like Heathrow, Tokyo Haneda and New York JFK, the right to land at a given time is a scarce, tradeable asset; Heathrow slot pairs have changed hands for tens of millions of dollars. An airline can want a route, have the airplane, and still wait years for the takeoff and landing times to exist, a bottleneck our guide to airport slots and curfews covers in depth.
| Factor | Question the planner asks | Typical deal-breaker |
|---|---|---|
| Demand | How many passengers, at what fare? | Thin market with low yields |
| Aircraft | Does the fleet have range and the right size? | No suitable gauge available |
| Rights | Does a treaty allow the flight? | Restrictive bilateral |
| Slots and gates | Can we land when passengers want to fly? | No commercially useful slot times |
| Competition | Who fights back, and how hard? | Incumbent adds capacity to defend |
| Connectivity | Does it feed the hub banks? | Arrival misses every connecting wave |
Step four: The clockwork of the hub
For a network carrier, no route is an island. A new arrival into the hub is designed to land just before a departure bank, so its passengers flow onward to dozens of destinations; the route's business case counts all that connecting revenue, not just the local traffic. This is why flight times sometimes look odd, a 6:05 a.m. departure exists because it catches the morning bank, not because anyone enjoys it. It is also why alliances and joint ventures matter so much to planners: a partner's hub multiplies the destinations a single new route can serve. Watch Atlanta on our live map around 8 a.m. Eastern and you can see the bank structure with your own eyes, or check which cities Delta's fleet is feeding right now through our Atlanta airport page.
Seasons, and the art of quitting
Networks breathe with the calendar. Mediterranean and Alaskan routes swell in summer; Caribbean flying peaks in winter. Planners publish seasonal schedules and shift aircraft between hemispheres like migrating birds. And sometimes the right decision is retreat. Airlines cut routes with the same cold math they use to launch them, redeploying the airplane to a better market, because the scarcest resource in the business is not money but aircraft time, and every hour a jet spends on a weak route is an hour stolen from a strong one.
Did you know?
When Qantas tested nonstop New York to Sydney flights in 2019 under Project Sunrise, the research flights carried scientists monitoring pilot brain waves and passenger melatonin to study how humans cope with 19 hours aloft. The route's business case depended as much on human endurance as on fuel burn, a reminder that route planning ultimately serves bodies, not spreadsheets. See how the science of body clocks works in our jet lag explainer.
When the bets go wrong
Competitive response is the wild card no model fully captures, and route planning is, above all, applied humility. Fuel prices double, currencies collapse, pandemics erase demand overnight, and a competitor can match your clever new route within a season and split the market that barely supported one airline. The A380 super-jumbo, ordered on forecasts of ever-growing hub congestion, met a world that preferred nonstop flights on smaller twins, and production ended after just 251 aircraft. Even the best-run networks carry a tail of experimental routes that will quietly disappear next year. The difference between good and bad planning departments is not that the good ones are always right; it is that they kill their mistakes faster.
Key takeaways
- Route decisions start with estimating hidden origin-destination demand, mostly traveling today via connections.
- QSI models forecast market share from schedule quality, but the forecasts are bets, not guarantees.
- Long-range twinjets like the 787 unlocked "long, thin" nonstop markets that jumbos could never serve.
- Treaties and airport slots can block a route no matter how strong the demand.
- Hub timing and belly cargo often decide a route's fate as much as passenger fares do.
- Good planners cut failing routes quickly; aircraft time is the scarcest asset an airline owns.
Frequently asked questions
How long does it take an airline to launch a new route?
Anywhere from a few months for a simple domestic route to several years for an international one requiring treaty rights, slots and new crew bases. Most routes are announced six to twelve months before the first flight to build bookings.
Why do airlines fly some routes only in summer?
Because demand is seasonal and aircraft are movable. A jet that serves Greek islands in July can fly ski markets or the Caribbean in January. Seasonal scheduling squeezes more revenue from the same fleet.
Why doesn't my city have a nonstop flight to a place lots of people visit?
Usually one of three reasons: the demand is thinner than it looks once fares are considered, no aircraft in the fleet matches the market's size and distance, or slots and treaty rights make the flight impossible or unprofitable at useful times.
Do airlines make money on every route they fly?
No. Networks deliberately carry some loss-making routes that feed profitable ones with connecting passengers, defend strategic markets, or honor slot-use rules. The network is judged as a whole, not flight by flight.
So the next time a new route appears from your home airport, read it as the end of a long argument someone finally won. And when a route quietly vanishes, know that somewhere a planner updated a model, sighed, and moved the airplane to a better bet, because in this business the map is never finished, only revised, one contested spreadsheet at a time.
