Measuring What the Passenger Feels
Lessons from building a directorate intelligence platform across a 22-airport estate — and the passenger product that feeds it.
Twenty-two airports, three customer-facing units, one Director accountable for the passenger experience across all of it. The first decision was the only one that mattered: the reporting layer and the working layer had to be the same layer.
Reporting after the fact is not measurement
In most institutions, performance data is assembled upward. The officer works in a notebook or a WhatsApp thread, a unit head compiles a monthly return, and by the time a number reaches the top it describes a memory — usually a generous one. Any dashboard built on that pipeline inherits both the lag and the optimism.
So in the platform we built for FAAN's Directorate of Public Affairs and Consumer Protection, the officer's task tool *is* the data source. A SERVICOM complaint moves across a five-stage board with an SLA clock on it; the resolution rate is a by-product of that movement, not a submission. A press inquiry is answered inside the response queue; the response-time metric is the queue's own history. A protocol service is ticketed, delivered and closed in the platform; patronage and revenue roll up automatically.
If your reporting layer is separate from your working layer, you are not measuring operations. You are collecting opinions about them.
A national average is usually a comfortable lie
An estate of 22 airports is not flat. Two gateways carry the passenger volume, the media exposure and the political attention; several regional stations handle a fraction of it. A score that treats them equally produces a number nobody can act on.
The fix is unglamorous: tier the registry, weight each airport by throughput, and weight unit contribution by proximity to the passenger — consumer protection carries the heaviest weight, because that is where the traveller actually experiences the Authority. Benchmarks come from ICAO reference standards rather than internal precedent, so "better than last quarter" cannot pass for "good."
Design the permissions like an org chart, not a CRUD table
Role-based access across a multi-site public estate is an organisational design question wearing an engineering costume. We mapped who is accountable for which decision — Director, unit head, station officer — and encoded that, instead of inheriting a generic admin/editor/viewer hierarchy nobody in the building recognises. A permissions model an officer can explain to a colleague in one sentence is the only kind that gets respected.
The passenger product is part of the instrument
The same directorate is accountable for services that are also revenue lines: protocol assistance, guided tours, terminal commerce. Selling those through channels that leave no record means the service promise and the commercial performance are both unmeasurable. So the passenger product — booking, officer matching, QR-verified confirmation, corporate wallets and approval flows — was built as the front door of the same accountability model, not as a separate marketing asset.
What transfers beyond aviation
Three lessons hold outside airports. Name the one hostile constraint honestly and let it set the architecture. Make the working tool the measurement instrument, or accept that your metrics are self-reported. And weight your aggregates the way reality is weighted, because an average that hides the two nodes that matter is worse than no number at all.
See the delivered platforms in the FAAN case study. If you are building under similar constraints, the Phase-0 GTM Scorecard will show whether your weakest link is the product or the distribution nobody has planned yet.