Planning for Participation:
Redesigning Airport Layout
How external AI assistants are reorganizing the earliest layers of travel planning — and what it means for airlines arriving second.
Role
Research Lead
Role
Research Lead
Role
Research Lead
Role
Research Lead
Role
Research Lead
Role
Research Lead
Summary
Problem
Design
Insight
Impact
“AI was not just another source in the funnel. It was increasingly the place where the planning problem itself was being framed.”
Summary
Airport departure lobbies are transitional spaces that have to absorb rising passenger volumes inside fixed footprints. As demand grows, the question isn't only how to add capacity — it's how to make the space already in use work better. I led research at ORD, a priority hub for redesign, to understand why congestion was forming where it was and how the lobby could support more throughput, segmentation, and self-service without proportional increases in staffing or footprint. The study combined day-of-travel app data, operational data, behavioral observation, layout analysis, and academic literature on the social psychology of uncertainty and the anthropology of transitory spaces. The central finding was that congestion was being created upstream — by uneven inflow, cross-traffic from non-departing passengers, signage that increased interpretation costs under load, and stations that customers couldn't easily access without backtracking. The reframe was that lobby throughput is a coordination problem, not just a capacity problem — a system co-created by the airline's design conditions and travelers' moment-by-moment interpretation of them. That framing now guides ORD's modernization, including a commitment to two-stage kiosk and bag-check throughout the lobby.
Problem Space
Air travel is entering a period defined by higher passenger volumes, tighter operational margins, and rising expectations for clarity and self-service — especially at hub airports, where small frictions compound quickly into systemwide delays.
For the departure lobby, that means redesigning environments to support greater volume with clearer service segmentation, stronger self-service, and more reliable throughput, without relying on proportional increases in staffing or footprint. ORD was selected as a priority hub for this work, and from a research standpoint it offered an unusual opportunity: to step back from incremental fixes and examine foundational elements of the day-of-travel experience — elements that are difficult to change at scale once built.
The Lobby team had translated United's broader strategic priorities into a set of concrete principles for how traffic should enter, distribute, and progress through the space — forward, distributed, consistent, flexible, self-service. Each principle described an intended pattern of movement and decision-making. The role of research was to evaluate the current-state ORD lobby against those principles from a customer experience perspective, and to identify where the existing environment supported the intended mode and where it broke down.
That made the central question more specific than "is this lobby crowded?" In security-adjacent, time-pressured contexts, customers have to act quickly with limited deliberation. When the environment clearly supports the intended pathways, traffic distributes as designed and throughput is sustained. When it doesn't, customers hesitate, backtrack, or concentrate in ways that undermine the system's performance. The research had to show where ORD's design supported the intended customer behavior, where it didn't, and what that meant for the redesign.
Approach
The work was conducted over a multi-week period and combined four elements.
Quantitative analysis of digital and operational data — including day-of-travel app usage (especially check-in and checked bags), traffic volumes across lobby entry points, kiosk and bag-drop utilization, and security throughput across checkpoints — to establish where demand concentrated, where capacity went underused, and how patterns varied by location and touchpoint.
On-site behavioral observation focused on entry-point choice, initial orientation, navigation, backtracking, queuing, and social herding behaviors — to document how travelers actually committed to pathways under time pressure, particularly near kiosks, bag drop, and security.
Layout and service-organization evaluation examining how entry points, signage, the placement and orientation of kiosks and counters, security checkpoint positioning, and circulation paths influenced movement patterns.
Literature review spanning customer experience, self-service adoption, service design, the social psychology of uncertainty and social proof in high-pressure settings, and the anthropology of transitory "non-places" — used to contextualize observed behavior and strengthen interpretation.
The design choice that mattered most was treating the lobby as a coordination system rather than as a sum of features. Operational data alone would have shown where congestion formed but not why. Observation alone would have shown what travelers did but not whether the pattern was structural. Pairing the two — and grounding both in literature on how people behave in uncertain, transitory environments — produced something more durable: an account of how design conditions and customer interpretation jointly produce the throughput patterns the lobby is trying to manage. That framing became the analytical spine of the work.
What We Found
Throughput Is Co-Created, Not Engineered
Anthropologically, airports have sometimes been treated as "non-places" — transient spaces organized around circulation and compliance rather than identity or relationship, with much discussion devoted to the apparent tension between efficiency and depersonalization.
Social and environmental psychology complicates that picture. When travelers enter the lobby, they effectively accept strict behavioral constraints in exchange for an environment that reduces functional decisions. They trade open-ended choice for a different kind of freedom: freedom from low-value decisions, so they can move confidently and direct attention toward more discretionary aims (food, comfort, time).
That dynamic isn't customer behavior layered on top of infrastructure. Lobby throughput is a value co-creation system: the airline shapes conditions through layout, signage, tool placement, and service design, while travelers complete the system through attention, interpretation, and follow-through. When those conditions support fast recognition and low-friction action, movement stays distributed and forward. When they don't, work shifts back onto travelers (second-guessing, backtracking) and onto agents (intervention, assistance).
This framing turned the Lobby team's principles into something evaluable. Instead of asking whether ORD had "good signage" or "enough kiosks," the research asked whether the space provided the conditions for participation — conditions that eliminate deliberation, make functional decisions low-effort and fast, and reserve choice for discretionary aims.
Inflow Patterns Were Concentrating Traffic Where It Couldn't Distribute
Traffic enters the ORD lobby from three sources: curbside drop-off along the terminal, central escalators from public transit and the airport tram, and north and south escalators from the level below.
Curbside traffic is disproportionately heavy on the north end, in front of the economy area. Vehicles queue most densely in front of "Check-In" (economy) signage, while traffic thins near Premier and Bag Drop Shortcut signage. This pattern suggested that curbside signage was shaping misleading expectations about eligibility. The juxtaposition of "Check-In" with "Premier" invited a false inference: that all south-end interior services were reserved for premium travelers.
Travelers arriving via central escalators enter directly into the middle of the lobby, with the majority flowing right toward economy kiosks and a standard security touchpoint. That flow meets the curbside traffic from the north end, creating congestion at the economy kiosks — exacerbated by the fact that the busiest security checkpoint intersects the kiosk area, creating a pinch point when security lines spill into the lobby.
A further factor was non-departing traffic. The central ride-hailing pickup area sits at the south end of the lobby. Travelers arriving at ORD who hail a ride after picking up bags must come upstairs and travel south. Those ascending via the north escalator must traverse the entire departure floor. The result: a steady stream of non-departing passengers moving laterally through active departure zones, contributing to congestion concentrated around economy check-in.
Kiosk and station usage mirrored these dynamics. Stations along main paths saw high utilization. Stations requiring lateral movement or backtracking went underused — even when they would have served the same function more efficiently.
Signage and Layout Increased Interpretation Costs Under Load
Given the inflow patterns, the lobby's ability to manage throughput depended heavily on the legibility of signage and the lobby's ability to absorb peak demand. ORD's central signage monument, while not poor, contained elements that increased complexity rather than reducing it.
Adjacent signage assumed prior actions ("already added your bags") that many travelers couldn't easily place in time or channel. Most consequentially, security nomenclature ("Pre-Check," "Security 1," "Security 1A") was functionally opaque, especially for first-time ORD travelers. Under time pressure, travelers were being asked to do interpretive work the environment should have been doing for them.
Distribution at ORD relied almost entirely on travelers self-selecting alternatives, and that approach didn't always work in practice. The Additional Services station illustrated the tension most clearly: its function was so narrowly defined that, instead of absorbing demand, it often redirected travelers back into the very queues they were trying to avoid.
The pattern across all three findings was the same. ORD wasn't failing at throughput because of capacity. It was failing because the environment was making travelers do interpretive work in moments where they were least able to deliberate — and that interpretive work was being absorbed by queues and agents, not by the design.
4. The system worked best when customers could manage changes early; friction grew around price, flexibility, and brand fit.
Assigned seating remained active after booking. Thirty-three percent of post-trip respondents changed or upgraded seating after the initial booking process. Among those who changed, outcomes were strongest when the change happened digitally: travelers who changed seats in the app or website were 91% satisfied overall, compared with 77% for those who changed seats at the gate.
Gate agents also played an important recovery role. Fifty-two percent of respondents spoke with a gate agent during the trip. Those interactions were often tied to carry-on/bin space, sitting with a travel group, seat assignment changes, moving to a different seat, or upgrade questions. Speaking with a gate agent was not itself a negative signal; customers who did so had a higher promoter rate than those who did not. The issue was what created the need for recovery in the first place.
What Changed
The research translated into three overarching evaluation principles now used as shared reference points across the Lobby team and architectural planning:
Include the full lobby within customers' range of experiences. Ensure travelers entering from the north end aren't effectively cut off from usable options or forced into long, conflicting cross-traffic just to access core functions.
Align functionality with behavior and expectations. Make sequencing and naming reflect how travelers actually interpret clearance levels and progression, rather than forcing interpretation under time pressure.
Minimize customer decisions. Reduce moments where travelers must stop, compare options, or infer what applies to them. Use landmarks and information structures that mirror the lobby's physical logic so orientation and selection happen quickly.
The findings have already informed concrete operational changes at ORD. Bag Drop Shortcut — the intermediate bag-drop option between Premier and Economy — has moved to curbside-exclusive, addressing the inflow concentration that the research identified. Security signage nomenclature has been redesigned to replace functionally opaque labels with naming that aligns with how travelers actually interpret progression.
More substantially, the work informed a commitment to two-stage kiosk and bag-check throughout the ORD lobby — a design that aims to enhance self-service, ease segmentation issues, and improve throughput by structurally aligning the space with the patterns the research surfaced. This is the largest customer-facing operational shift the work has driven, and it reflects the central finding: throughput improves not by adding capacity but by making participation easier under load.
The research was presented to senior leadership across Lobby, Airport Experience, Customer Strategy & Insights, Real Estate, and Customer Experience. The framework is now used as evaluation criteria for lobby concepts at ORD and beyond — providing a repeatable way to evaluate proposed layouts not by feature count but by how well they reduce functional decision burden in the moments travelers are least able to deliberate.
The deeper takeaway, both for hub design and for service environments more generally, is that throughput is a coordination problem before it's a capacity problem. The leverage is rarely in adding more — more counters, more signs, more agents. It's in making the system the customer is asked to participate in legible enough that the participation happens fluently, on its own.