Ask anyone who has worked inside an airline whether the commercial and operational sides of the business are well coordinated, and you will get a polite answer. Ask them whether a decision made in network planning has ever created a problem that operations absorbed without anyone connecting the two, and you will get a different kind of answer.
The misalignment between airline functions is not a secret. It is one of those problems that is widely understood, frequently discussed in leadership offsites, and almost never solved. Not because the people involved are not trying. Because the structural conditions that produce misalignment are built into how airlines are organized, how decisions are made, and how performance is measured.
Understanding those conditions, and what it would take to change them, is more useful than another call for better cross-functional communication.
What misalignment actually looks like
It rarely looks like conflict. That is what makes it hard to see and harder to fix. Misalignment in an airline looks like a network planning team scheduling a flight into an airport with an arrival time that is legally within the curfew window, but only barely. On a normal day the flight operates fine. On a day with a minor upstream delay, a late inbound aircraft, a slow gate push, a thirty-minute ATC ground stop, the flight can no longer legally arrive before the airport closes. It diverts. It cancels. Passengers are stranded, crew are out of position, and operations spends the next twelve hours managing the consequences.
Revenue management had no visibility into the fragility of that arrival window. They priced and yielded the flight normally, correctly, given what they knew. The network planning team scheduled the flight legally, correctly, given the variable cost model they were working from. Neither team made a bad decision. The system made a bad decision, because no single part of it had the full picture.
This example exposes something important about how operational risk is treated in network planning. Demand is modeled quantitatively, forecast curves, booking pace, yield elasticity. Variable costs are modeled quantitatively, fuel burn, per-cycle reserves, landing fees. Operational risk is evaluated qualitatively. Someone looks at the curfew window and decides it looks workable. Nobody runs a model that says this arrival time has a 23% probability of causing a diversion based on historical delay patterns at this station, on this aircraft type, during this season, even though that number is calculable from data the airline already has. The qualitative judgment gets the flight scheduled. The quantitative consequences show up later in operations.
It looks like a marketing team building a campaign to drive demand to a specific flight, investing budget, targeting the right audience, generating genuine interest. At the same time, revenue management looks at the booking pace, sees demand coming in, and reduces price to accelerate load factor. Two teams both doing their jobs correctly, working in opposite directions, with neither aware of what the other is doing. The marketing spend generates demand that revenue management then discounts away.
It looks like operations absorbing a schedule that network planning designed without full visibility into what that schedule does to crew resources at a specific base, on a specific day of the week, during a specific season. The schedule passes the variable cost threshold. It is legal. It is flyable. It is also brutally difficult to operate, and the people who feel that difficulty are the ops team, not the planning team that designed it.
These are not failures of individual judgment. They are failures of information architecture. The right people are making the right decisions with the information they have. The problem is that the information they have is incomplete in ways they cannot fully see from where they sit.
Why operations absorbs the pain
When alignment breaks down, the impact does not distribute evenly across the organization. Network planning may receive criticism eventually, after the schedule has operated, after the results are in, after someone does the analysis that connects the planning decision to the operational outcome. But that criticism arrives months later, filtered through layers of post-hoc rationalization, and is rarely specific enough to change how the next planning cycle works.
Revenue management sees the financial outcome in their reporting. If a flight underperforms, there are reasons, demand was softer than forecast, a competitor did something unexpected, weather affected the market. The contribution of a planning assumption to a revenue outcome is difficult to isolate after the fact.
Operations feels it immediately, specifically, and personally. The flight that was designed with a tight turn at a congested hub is not an abstraction to the ground crew working it. The pairing that looks efficient in the crew planning system but creates a brutal overnight sequence is not a spreadsheet entry to the crew flying it. The station that was nominally staffed for the flying level but is actually running on overtime because the schedule structure was not designed with their utilization in mind is a daily reality for the people managing it.
Operations is the ultimate downstream recipient of every decision made upstream. When those upstream decisions are made without full visibility into their operational implications, operations absorbs the consequences. This is not a complaint about any particular function. It is a structural feature of how airlines are organized, and it shapes the incentives of every team in the system.
There is a paradox buried in this dynamic that deserves to be named directly. The decisions that create operational instability are almost always made in the pursuit of financial efficiency. A tight turn reduces ground time and improves aircraft utilization. A lean station staffing level reduces labor cost. A curfew-adjacent arrival time maximizes the usable flying day. Each decision looks efficient in isolation, evaluated against the information available at the point it was made.
But efficiency built on incomplete information produces instability. And instability is expensive in ways that never appear in the original model. A tight turn that misses its window generates a delay that propagates through the rest of the aircraft's flying day. A lean station that cannot absorb a disruption generates overtime, missed connections, and customer recovery costs. A curfew-adjacent arrival that diverts generates IROPS costs that can exceed the revenue of the entire flight. The efficiency gain that justified the decision is wiped out many times over by the instability it created.
The pursuit of financial efficiency with imperfect information does not just fail to achieve its goal. It actively produces the opposite outcome. That is the paradox that sits at the center of the airline alignment problem, and it is why solving the information problem is not just operationally important. It is financially essential.
The coordination mechanisms that exist, and why they are not enough
Most airlines have mechanisms designed to address this. Cross-functional planning reviews. Schedule evaluation processes where operational teams provide feedback on draft schedules. Revenue and operations meetings where commercial and ops leadership align on priorities.
These mechanisms help. They are not sufficient. The fundamental limitation is speed. A cross-functional planning review happens on a schedule that is measured in weeks or months. The decisions it is meant to coordinate happen continuously, in real time, across dozens of functions simultaneously. By the time a review catches a coordination problem, the decision that created it has often already been made, or the window for changing it without significant disruption has closed.
The feedback loop is also incomplete. A network planning team that receives operational feedback on a draft schedule gets a point-in-time view of the operational implications of a specific schedule configuration. What they do not get is a dynamic model of how those implications would change if they modified the schedule, if they shifted a frequency, changed a departure time, swapped an aircraft type. The feedback is evaluative rather than generative. It tells you what is wrong with the current plan. It does not show you what a better plan would look like.
There is a version of this problem that has been partially solved. At one airline, the commercial and operational planning functions were connected through a suite of integrated models, capacity plans generated schedules, schedules generated crew requirements and revenue forecasts, and the entire system could be run through a Monte Carlo simulation to evaluate operational performance across thousands of scenarios. Multiple iterations could be evaluated automatically, in a fraction of the time that manual review would require.
The result was something close to what good alignment should look like: a planning process where the downstream operational and financial implications of a scheduling decision were visible at the point the decision was being made, not weeks later after the schedule was already being sold. That kind of capability is not common. It is the direction the industry needs to move.
The measurement problem underneath the alignment problem
Underneath the coordination failures is a measurement problem that rarely gets discussed directly. Each function in an airline is measured on metrics that are specific to that function. Network planning is measured on network contribution, market share, and schedule quality metrics. Revenue management is measured on yield, load factor, and revenue per ASM. Operations is measured on on-time performance, completion factor, and cost per available seat mile. Marketing is measured on campaign performance, brand metrics, and demand generation.
These metrics are not wrong. They are appropriate measures of functional performance. The problem is that they are not connected to each other in a way that reflects the actual interdependencies of the business.
When revenue management drops price on a flight that marketing is actively promoting, neither team's metrics capture the combined cost of that coordination failure. Revenue management sees load factor improve. Marketing sees demand generated. The fact that the marketing investment and the revenue discount worked against each other is visible only in an analysis that connects both sets of data, an analysis that no one has an obvious incentive to do, because it would require attributing a performance problem to two teams simultaneously.
When network planning adds a frequency that creates operational pressure that ops absorbs through overtime and delay, network planning sees a route contribution that meets their threshold. Operations sees cost and performance metrics that look worse than they should. The connection between the two is visible in retrospect, if anyone looks for it.
The incentive to look for these connections is weak when each function is rewarded for its own metrics. The visibility to find them is limited when the data lives in separate systems that were not designed to be analyzed together.
What changes with genuine integration
The alignment problem does not have a simple organizational fix. Restructuring reporting lines, adding coordination roles, increasing meeting frequency, these interventions address the symptom rather than the cause.
The cause is that the information required to make well-coordinated decisions does not flow to the people making those decisions at the speed and granularity required to act on it. Fixing that requires building the analytical infrastructure that makes the full impact of decisions visible before they are made, not after.
This is what AI makes possible in a specific and practical way. A revenue management system that has visibility into the marketing calendar does not drop price on a flight that is already being promoted. It coordinates the timing of its own interventions with what marketing is doing, rather than working against it without knowing.
A network planning process that can model the full operational implications of a frequency decision, not just the variable cost contribution, but the crew resource impact, the station utilization implications, the connection revenue effects, makes different decisions than one working from a partial model. The problems that operations would otherwise absorb are caught at the point of planning rather than at the point of operation.
A cross-functional view of the schedule that updates dynamically as conditions change, as demand evolves, as operational constraints shift, as competitive dynamics move, gives leadership a real-time picture of where the plan is under stress and where it is performing, rather than a static view that is already out of date by the time it is produced. None of this is hypothetical. The analytical capability to build these connections exists. The organizations that build them into their planning and decision-making processes will make better decisions, coordinate more effectively, and spend less time managing the downstream consequences of choices made without complete information.
The alignment problem is a real problem. It has a tractable solution. The gap between where most airlines are and where they could be is largely a question of whether building that solution is treated as the priority it deserves to be.