BCG Says Airports Need an AI Nervous System. They're Right — But They Forgot About the Tummy

BCG published a piece last month called “The AI Nervous System Guiding Leading Airports.” It’s a solid read. The thesis: airports that build integrated, AI-driven operations control centers — what BCG calls a “digital nervous system” — are outperforming their peers on throughput, reliability, and commercial outcomes. The airports that don’t are burning money on disconnected pilot projects that never scale.

They’re right. And if you run an airport, operate concessions, or own a brand inside a terminal, this article should be required reading. But here’s my read on it: BCG nailed the diagnosis on the operations side and barely touched the part of the airport that actually generates the non-aeronautical revenue everyone’s chasing.

What BCG Got Right

The core framework is sound. Most airports pursuing AI are doing it wrong — long lists of use cases, no shared data layer, siloed stakeholders, and no clear governance for real-time decisions. BCG calls this out directly, and they’re not being diplomatic about it. The result is a bunch of proof-of-concepts that prove the concept and then die in a drawer. Part of that is a structural procurement issue. Different functional areas look for service providers that meet their specific needs. The “digital nervous system” concept requires multiple departments to have data needs met by providers of varying sophistication levels who do radically different things. The other problem is that the speed at which Artificial Intelligence can do new things is greater than the speed at which large organizations with state procurement guidelines can even answer the question. So what airports have successfully launched implementations?

The airports that are winning — BCG cites a Nordic operator, a major European hub, and an Asia-Pacific facility — are doing something different. They’re building a shared operational picture across airlines, ground handlers, customs, and facilities management. They’re using predictive intelligence to anticipate congestion and disruptions before they cascade. And they’re putting decision rights in a single control center where stakeholders actually resolve conflicts instead of protecting their own turf.

The numbers BCG references are striking. One Asia-Pacific airport’s AI transformation could generate an estimated $380 million in direct benefits between 2030 and 2040, with $4.8 billion flowing to airlines and ecosystem partners. That’s not incremental. That’s infrastructure-grade value creation without building a single new runway.

What They Missed

Here’s where I part ways with the consulting framework. BCG identifies four value pools: aeronautical revenue, non-aeronautical revenue, reduced operating expenses, and deferred capital expenses. Non-aeronautical revenue gets a mention. But the article treats concessions — the restaurants, shops, and services that passengers actually interact with — as an afterthought. A line about “pop-up shops” and “retail offers.” That’s it.

This is a blind spot, and it’s a big one.

Non-aeronautical revenue isn’t a nice-to-have line item. At most major U.S. airports, concessions revenue is a primary driver of the commercial program. We’re talking about programs worth billions (in some cases tens of billions) of dollars over a contract term. The concessions ecosystem — operators, brands, developers, ACDBE partners — is where the passenger experience actually lives. Nobody’s airport satisfaction survey is about how smoothly the baggage belt allocation algorithm performed. They remember whether they got a good meal, whether the line moved, and whether the experience felt like it belonged in that city.

BCG’s “digital nervous system” is powerful precisely because it could transform that experience. But you have to design for it intentionally — it won’t happen as a byproduct of better gate assignments.

The Concessions Gap

Think about what an integrated data layer could actually do for the commercial side of an airport:

Real-time passenger flow data could tell a concessionaire exactly when their rush is coming — not based on a static flight schedule, but based on actual security throughput, gate assignments, and connection timing. A restaurant in Terminal C could know 30 minutes in advance that 400 passengers from a delayed international arrival are about to clear customs and walk past their door. That’s not a nice dashboard. That’s the difference between prepping for the rush and watching customers walk past a closed kitchen.

Predictive dwell time modeling could reshape how airports think about retail placement, lease structures, and even which brands belong in which locations. If you know that passengers on certain routes consistently have 90+ minutes of dwell time and skew toward premium spending, you can curate a concessions mix that actually matches the traveler — not just fill square footage.

We’re already seeing airports move this direction. DFW’s latest RFP explicitly requires bidders to offer self-checkout, participate in mobile ordering platforms, and integrate with the airport’s tech ecosystem. They’ve built traveler personas and are asking concessionaires to match their offerings to specific segments. That’s the commercial side reaching for exactly the kind of integration BCG describes — but from the bottom up, through procurement, not from the top down through an operations control center.

The Maryland Aviation Administration has an open solicitation for integrated airport data, display, and management systems that starts to connect these dots. The airports that figure out how to link operational intelligence to commercial execution — not just ops to ops — are the ones that will capture the full value BCG is describing.

What This Means for Operators and Brands

If you’re a concessions operator reading the BCG article and thinking “this is an airport authority problem, not mine” — reconsider. The airports building these systems are going to start requiring interoperability from their concessionaires. Open APIs, data-sharing commitments, real-time reporting. The RFPs are already moving this direction.

The operators who get ahead of this will have a meaningful edge. Not because they’ll build their own AI nervous system, but because they’ll be ready to plug into one. That means investing in tech infrastructure that’s flexible, being willing to share anonymized transaction data, and framing your RFP responses around passenger outcomes — not just rent tiers and revenue projections.

For brands exploring airport entry, the implication is similar. The days of winning a concessions unit purely on brand recognition and a strong pro forma are fading. Airports want to know how your concept fits the traveler profile at a specific gate, how your ordering technology reduces dwell time friction, and whether your systems can talk to theirs. Gong cha’s kiosk rollout(I talk about it here)  — 90% of transactions moving from counter to kiosk, higher average check, higher satisfaction — is the kind of operational transformation story that wins evaluations now.

The Bottom Line

BCG is right that airports need a digital nervous system. But a nervous system that only governs operations is like a body that can move its legs but can’t taste, smell, or feel. The commercial experience — the concessions program — is where passengers form their impression of an airport. It’s where non-aeronautical revenue lives. And it’s where the next generation of AI integration will separate the leaders from everyone else.

The airports and operators that treat technology as a passenger experience strategy, not just an operations optimization tool, will be the ones capturing the full value. That’s the thesis we’ve been building around at Clearport, and it’s encouraging to see BCG validating the infrastructure layer that makes it possible.

If your organization is thinking about how AI and digital integration fits into your concessions strategy — whether you’re an airport authority structuring your next RFP, an operator preparing a response, or a brand evaluating airport entry — I’d welcome the conversation. Schedule a call here.

Faraji Robinson is the founder of Clearport, an AI-native airport concessions advisory firm based in Washington, D.C. He advises operators and brands on RFP strategy, deal structure, and market entry across major U.S. airports.