Tinder & the Last Mile

The “last mile” has become an increasingly popular buzzword as transportation systems have grown faster and more efficient. Consider the following examples of the Last Mile Problem:

  • FedEx and UPS combined to rack up $2.8 million in New York City parking tickets during the first quarter of 2013, accepting those fines as a cost of doing business.
  • Reaching most US airports from their corresponding cities is a notoriously painful process that adds significantly to the cost of air travel, in the form of a long transit trip or an expensive cab ride.
  • Amazon’s fastest broadly available shipping option still only guarantees delivery the following day, although the merchandise rarely spends more than a few hours moving between its origin and destination cities.

In all of these cases, an efficient transportation system is bedeviled by its relationship to the outside world—by the origins and destinations where the actual people are located. The passengers, inconveniently, don’t want to live as close as possible to the infrastructure of high-speed travel, nor can they, mostly. The air transportation network and the FedEx supply chain operate as closed systems that move people or goods smoothly between their hubs and spokes but the final legs of those trips, on city streets lined with homes and offices, are far more expensive than the long hauls. In the case of air travel, the carriers avoid the problem altogether, leaving it for others to solve—they wouldn’t be profitable if they took passengers home from the airport. The traveler’s jarring transition from the airside to the groundside accounts for much of the inconvenience associated with flying. FedEx, on the other hand, is stuck with the last mile, hence their millions in parking tickets and consequent loss in profit margin.


The Last Mile Problem arises at the interface between two systems: the messy, irrational human environment and the rationalized networks that are optimized to achieve a few goals as efficiently as possible. I’ve written before about how those two systems find themselves at odds in the modern world, with the human city actually getting in the way of the “machine city” that comprises airports, seaports, bridges, and freeways, and making it harder for those networks to accomplish their goals. For infrastructure ostensibly built to serve people, the people themselves pose a surprising obstacle. In a broad sense, the Last Mile Problem is the problem of humans being humans and not machines.

Mankind has managed to annihilate distance quite successfully since ancient history, and the Last Mile Problem didn’t exist until the last mile wasn’t the only mile that most people had to think about. The last mile, therefore, is relative: The distance that we define as the last mile becomes shorter and more frustrating as we overcome the physical distance between ourselves and the things we want through faster travel across greater distance. Software has seemingly solved the Last Mile Problem in major cities by enabling the on-demand economy to flourish and become more affordable, yet we call the last mile a problem more frequently now than ever before, because our expectations have risen just as fast and the last mile is now measured in minutes. Waiting a half hour for a car service is deemed intolerable for those accustomed to getting picked up in 3 minutes via an app, although the former was a luxury not long ago and still is for most people. Likewise, the ability to order any product online and have it delivered the following day, unthinkable even a decade ago, now represents another problem: A day is too much, 24 hours is the new last mile, and the challenge of guaranteeing same-day delivery is a problem that needs to be solved.

The smartphone revolution, and the app ecosystem that has since emerged, suggest that a version of the Last Mile Problem extends beyond the transportation domain. Greg Lindsay, in his recent exploration of mobile dating apps and their effect on urban social life, asks whether apps like Tinder and Happn are augmenting face-to-face interactions or merely competing with them. Put another way, can we live well inside and outside of these “systems” simultaneously, or must we gradually abandon one for the other?

An app like Tinder may or may not improve our lives, but it certainly makes one aspect of dating more efficient, letting us sort through 100 potential mates in the time we’d spend talking to one or two people at a bar. Like the Acela Express, or many other transportation systems, Tinder achieves its efficiency by forcing its users to adhere to the constraints of the platform (the Acela train makes very few stops; Tinder requires you to swipe right or left instead of miring yourself in time-consuming, ambiguous signaling). By externalizing the steps in the process that they don’t streamline, each platform creates its own Last Mile Problem. The Acela will get you to Union Station in DC quickly, but you’ll be stuck in traffic taking a cab the rest of the way to your house in suburban Maryland. Similarly, Tinder will match you with plenty of people who also liked your profile photo, but the rest—the last mile—is up to you, and from that point onward, the process is often no better than what it replaced.

The last mile, then, is always located where machines leave off and humans are forced to take over. Unfortunately, the more time we spend in the controlled system, the more difficult the transition, since the rules of the platform require us to behave more like machines than humans, mindlessly swiping and clicking or passively riding before suddenly needing to reawaken all the faculties that have fallen asleep. It takes entirely different skills to navigate these platforms and systems. Marshall McLuhan called every new medium an extension of man, but also recognized that each extension also involved an amputation of what it replaced. Few will argue that the smartphone hasn’t “amputated” certain mental capacities, or at least allowed them to atrophy.

Lindsay’s examination of dating apps concludes by acknowledging that the most valuable human interactions still occur in person, regardless of which apps or technologies helped to facilitate those interactions. In other words, the last mile, however defined, is still where the most important human affairs happen (and for millennia, the last mile was where everything happened). Yes, the last mile is a frustrating problem in transportation, but the space where we traverse that last mile is the space we actually inhabit, and it’s worth protecting from the various engineered systems with which it competes for our attention and resources. Whether a measure of distance, time, or social interaction, a last mile will always exist where our technological extensions end and our embodied selves begin. As that technology advances and the relative length of the last mile recedes, approaching invisibility, we humans will finally find that the last mile is just…us.

2 Comments on “Tinder & the Last Mile”

  1. markbao says:

    Excellent post and a great ending. Two things:

    1) It seems to me that in any given system, the last mile is an artifact of disproportionate optimization. That is, in the case of Tinder, the entirety of the system is the dating process. Tinder optimized the matching process, but now matching is easier and disproportionally so to the rest of the dating process (which is still just as difficult). Acela optimized city-to-city transit, but disproportionally so to the entire “my house to the hotel” process, so the lagging part is the last mile. It just so happens that in many systems, the point where they stop is the human part, because the technology isn’t there to change that human aspect.

    2) What is this human aspect that divides the last mile? Maybe it’s expectation. Five years ago I was fully content getting an Amazon package in 5 days. Now, some days I lament how long 2-day shipping takes. While these optimizations in e.g. shipping are objectively improvements, I feel like the subjective feeling of that optimization resets quickly so that we come to expect that as the standard. 30 minutes for a cab is crazy when we get an Uber in 3, as you say. The last mile, I think, has roots in the quickly-resetting expectations of humans in the system. And (risking taking this a bit far but let’s see where it goes) it’s not because humans are hard to work with as static agents in that system, it’s because they seem almost adversarial to any optimization, in that they accept it, see that new last mile as a baseline, and then come to expect more.

  2. Ben says:

    This guy is already trying to automate away some of the Tinder Last Mile: http://crockpotveggies.com/2015/02/09/automating-tinder-with-eigenfaces.html

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s