PaxInsight: Real-Time Passenger Satisfaction Intelligence

The problem with airline surveys

The airline industry measures everything. Fuel burn per seat-mile, on-time departure rates, load factors, ancillary revenue per passenger. Operations centres track engine degradation in real time and can pinpoint the cargo hold temperature to a tenth of a degree. Yet for all this precision, most airlines cannot reliably answer a simpler question: are our passengers satisfied?

The reason is methodological. In-flight surveys reach only the motivated minority, the very satisfied and the very angry. Business-class passengers respond at different rates from economy travellers. Responses arrive weeks after the flight, by which time operational context has been lost. A carrier might conclude that a mechanical delay caused a satisfaction collapse, when the real problem was that its sample over-represented disgruntled passengers. The survey measured complaint propensity, not experience.

PaxInsight was built to fix this. Developed by IATA and operated across a global network of carriers, the platform tackles a basic asymmetry: airlines spend billions on seats, catering and lounges, yet invest almost nothing in understanding whether those investments actually shift passenger perception in any measurable, comparable way.

How the sampling works

Each month, IATA generates a sampling plan using airline schedule data. The plan ensures that all origin-and-destination (O&D) pairs within a given region are represented proportionally. For airlines affiliated with IATA's Direct Data Solutions (DDS), ticket numbers are selected directly through the DDS system. For non-DDS airlines, IATA draws a random sample from passenger data provided by the carrier. Either way, the passenger receives the same standardised survey invitation, delivered by email 24 to 48 hours after landing.

The minimum target is 150 responses per airline, per region, per cabin class, per month. Response rates average 8 to 12 percent, comfortably above industry norms. Because the sampling is proportional to actual passenger volumes on each route, a large hub like London Heathrow to New York JFK will contribute more responses than a thinner route, but both are represented at their correct weight. The result is a dataset that genuinely reflects the airline's passenger base rather than the subset of travellers who felt strongly enough to volunteer feedback.

Survey architecture: 60+ attributes across the full journey

Rather than a single satisfaction question, PaxInsight captures feedback across more than 60 travel attributes, organised into three phases. Pre-flight covers booking, check-in, the departure lounge, boarding, and any transfer experience at the origin airport. In-flight covers cabin environment, seat comfort, cabin crew, cleanliness, Wi-Fi and connectivity, entertainment, and food and beverage. Post-flight covers arrival experience and transfer at the destination. Each attribute is scored on a 1-to-5 scale, and the overall Net Promoter Score is collected separately.

Branching logic keeps the survey manageable. Economy passengers on short-haul flights skip lounge questions entirely. Long-haul business travellers receive additional items on sleep quality and lie-flat seat comfort. Connecting passengers see a dedicated transfer segment. The result is a survey tailored to each journey type, with completion rates exceeding 94 percent and average completion times of 9 to 12 minutes.

One client found that business-class satisfaction was driven almost entirely by service consistency (crew attentiveness, meal timing) rather than by the seat product or champagne selection. Another discovered that economy passengers on short-haul flights prioritised boarding speed over in-flight entertainment. These are the kinds of insight that a generic "how was your flight?" survey cannot produce. They require deliberately designed questions targeting specific moments in the journey.

Interactive survey demo

Below is a condensed 12-question version of a PaxInsight survey. The actual surveys run to 25 to 35 questions depending on cabin class and route, covering 60+ attributes with extensive branching logic. This demo captures the core question types and flow. Click through to experience the methodology firsthand.

Dashboards: from responses to executive intelligence

Survey responses feed into Power BI dashboards that update monthly. The actual client dashboards included an executive summary, touchpoint deep-dives, pre-flight and in-flight breakdowns, airport-level analysis, and a key-drivers screen. Below are four representative visualisations showing the type of competitive intelligence delivered to airlines.

Platform scale and quality (typical monthly cycle)

These figures reflect the general operating parameters of the PaxInsight platform during the project period.

Impact: what the data changed

Over 12 months, I delivered monthly intelligence packages to three carriers. Each included an 8- to 12-page executive report, live Power BI dashboards with full drill-down, and ad-hoc analysis turned around within 48 hours. Below are four representative examples of how the data influenced business decisions.

The work showed that rigorously collected passenger experience data, stratified, timed and benchmarked, can inform decisions that airlines previously made on intuition. Each outcome above was supported by specific data points from the monthly reporting pipeline I built and maintained throughout the project.