Prologue: I'm going to mainly reference the airline industry as a case study to the question posed in this article. The ideas presented can hopefully be thought of as industry agnostic, however.
PNR = Passenger Name Record
If you've ever booked a flight, your booking probably came with a 6-character alphanumeric record locator, better known as a PNR, or passenger name record. That random string of letters and numbers is both temporary and confidential. It's a password of sorts that lets you access your reservation online and at the airport.
Let me pose a question:
If you had to rebuild the infrastructure that powers airlines today, would you introduce the concept of a PNR?
The Challenge of Legacy
Airlines, not unlike many service-based industries, have been around for quite some time. On the other hand, the technology that underpins the way we as a society function has advanced materially. We are often digital-first, accustomed to convenience, tailored interactions, and seamless handoffs between technology and the world itself.
To be clear, this not to say that airlines haven't adapted to modern comforts we've afforded ourselves - they have! Frequent flyer programs, online reservations, manage my booking, digital tickets, etc. Yet they have indefinitely been bound by limitation and often constructs of previous systems and iterations before them. There's a reason many industries still run a mainframe computer in their headquarters.
Deepening this challenge is often that as an industry you have shared or common standards that exist for the benefit of all entities. This means even as an individual business, you may not be at the liberty to depreciate concepts that are heavily intertwined in downstream systems or processes.
Decoupling Consumer vs Enterprise Value
This brings me back to my original question. The last time you booked an Uber, whether it was ad-hoc or a scheduled ride, you probably weren't prompted or even provided a PNR. I'm almost positive one still exists, but as a consumer, having a unique identifying string that tracks the relationship between you, your driver, and the ride your on - provides no discernable value. Internally to Uber, I imagine this figure is a primary key in many operational and analytics systems.
So what's the difference?
When building a ride sharing platform from the ground up, with no incumbents, no existing infrastructure, and no integrations, Uber could be judicious about the data concepts it introduced for its platform. It wasn't riddled by the baggage of technical debt, and appears to have optimized for experience.
In this manner, technology fades away. The pipes and wires that connect riders to drivers millions of times a day feels as seamless, fluid, and natural.
Airlines have this magic too. The logistics to plan and operate thousands of flights per day, including planning the network, selling seats optimally, allocating crew, staffing airport operations, maintenance, regulatory, and more — it's a elegant dance that few consumers understand.
Building for the future
I'm not the first one to suggest graduating beyond legacy constructs that may not make sense in todays world, and I understand that change is often slow and gradual. Particularly when there are no off days for operations, and maintaining the effectiveness of production systems requires careful planning and execution. I'm encouraged by industry initiatives that seek to move the airline industry forward in one way or another, and believe it reaffirms this concept of viewing technology as an enabler rather than set of tools.
It allows you to reshape problems is if there were no rules. Even if the end state is not possible, charting a course towards ideal is better than simply planning for "next".
To quote one of my favorite articles recently, Extreme questions to generate new, better ideas: I hope the above thought experiment "jostles you out of tiny thinking".