Throughout my career in aviation, I've seen a lot of airlines launch modernisation initiatives. Most of them start the process by wanting to replace the core systems.
In many cases, the passenger service system, departure control system or crew and ops tools were built decades ago. They're expensive to maintain, hard to integrate and difficult to hire for. Swapping them for something modern feels like the obvious choice.
In reality, the airlines I see making the most progress take a different route. They keep their core systems running and start building around them. Their customers and staff notice the difference within months, while full replacement initiatives elsewhere are still in the requirements phase.
A lot of investment. Not always a lot of progress.
Almost 90% of airlines plan to invest more in passenger service systems in the next 12 months, according to BCG’s 2025 Turbulence to Transformation report. At the same time, IATA data shows that 80% of airlines identify legacy IT as a significant operational barrier, up from 65% in 2019. Most of that investment is going toward the very systems holding them back.
That cycle has consequences. When the CrowdStrike outage hit in July 2024, an American legacy carrier cancelled more than 7,000 flights over five days at a cost of around $550 million, according to their own SEC filing. Recovery took significantly longer than at other carriers affected by the same outage. Microsoft’s attorneys later pointed to the pace of IT modernisation as a contributing factor.
That was just one example. But I’ve seen versions of this across dozens of carriers. Replacement programmes that were going to solve everything, still in requirements phase two years after launch. And alongside them, carriers that started smaller, built around what they already had, and were already delivering results their customers could feel.
Why replacement programmes struggle – and what works instead
An airline can’t stop flying while it swaps out its passenger service system or departure control platform. Every major migration becomes a long exercise in running two worlds at the same time, at significant cost, for benefits that are months or years away. Leadership changes. Priorities shift. The programme that was going to solve everything becomes the programme nobody wants to own.
There’s also a timing problem that gets consistently underestimated. Any programme that takes three to five years to deliver results is exposed to everything that happens in that window. COVID dismantled roadmaps across the industry in weeks. In my experience, boards are patient up to a point. A multi-year programme with nothing visible to show is always vulnerable to the next disruption, the next leadership change or simply the next budget cycle.
Read article: What Are Our Accelerators and Why Do Aviation Companies Use Them
The carriers making consistent progress however, treat their core platforms as a stable foundation and put their energy into what sits on top. In practice, that means building a modern integration layer that can read from and write to legacy systems, then using it to launch new capabilities one at a time. Each one pays for itself. Each one gives the board something concrete to point to. Over time, the modern layer takes on more of the work and the legacy one naturally shrinks, as less and less of the operation depends on it.
Full service vs. low cost: same principle, different priorities
The approach can look slightly different depending on the business model, but the principle holds across carrier types.
Full service carriers tend to carry the heaviest legacy estates, with decades of customisation and dense alliance connections. For them, the integration layer is a way to deliver personalised, retail-style experiences without unpicking that complexity. The commercial case is usually found in ancillary revenue, direct channel conversion and NDC adoption, where the distribution opportunity has grown significantly in recent years.
A low cost carrier might start with simpler systems but faces tighter budgets and smaller teams. The priority is speed to revenue: ancillaries, direct sales and self-service that keeps call centre costs down. A LCC that builds automated rebooking on top of its existing DCS gets a result its customers notice within weeks. That result funds the next investment.
Neither group is waiting for a perfect platform before improving what customers and staff experience today.
Five steps to start with
1. Start with the operational problem, not the oldest system. Find the problem that costs the most in customer satisfaction, staff hours or revenue. For many carriers, disruption handling is the answer. For others, it's check-in queue management, lounge access control or crew scheduling inefficiency. The starting point should be something that, if solved, would be noticed by someone outside the IT team within 90 days.
2. Build the integration layer before building the capability. The integration layer isn't the exciting part of the project. It's the part that determines whether every subsequent project takes weeks or months. Done well, it becomes the infrastructure that every future capability runs on. Done poorly, every new project starts from scratch.
3. Choose the first use case for speed, not scope. The first project should be small enough to deliver in weeks and valuable enough to be noticed. This is not the time to solve everything. It is the time to prove that the approach works, build internal credibility and give the programme something concrete to show a board that has probably seen failed initiatives before.
4. Instrument everything from day one. Every capability should produce measurable output from the moment it goes live and those measurements should be operational rather than project-based. How many disrupted passengers were re-accommodated per hour. How check-in processing time changed. How many lounge access errors occurred that week. Those are the numbers that make the next funding conversation straightforward.
5. Let the legacy shrink on its own timeline. As the modern layer takes on more of the work, dependencies on the legacy platform reduce naturally. The decision to retire or replace the old system should come from the data on what still depends on it, not from a strategic assumption made at the start of the programme.
The board backs what keeps delivering
BCG's analysis found that between 2025 and 2027, the value airlines generate from digital technologies is likely to more than quadruple. This is an enormous opportunity. But it won't be captured by programmes that promise everything in three years and deliver nothing in the first twelve months.
The airlines doing this well choose to stop waiting for the right moment to replace everything and start building on top of what exists today. Each step produces something a board can see. That is what keeps the programme funded.
So if you're thinking of doing this, my suggestion is don't rush. Follow what works and take it one step at a time.
Or, get in touch.
