Uproar over Uber's wait-time charge reveals the pitfalls of designs that overpromise

Yu Siang is a star contributor for Tech in Asia and publishes exclusive, high-value content that serves the Asian tech community. Read more from star contributors here.

On October 3, Uber Singapore announced that it will be imposing a wait-time charge on riders who are more than three minutes late for their rides.

Wait-time charges are not a new concept; Grab, Uber's main competitor in Singapore, has been charging this for some time. (To be fair, Grab only starts imposing the late charge after five minutes, and its fees aren't subject to surge pricing, unlike Uber's.) However, this recent Uber announcement caused an uproar.

Most comments on Uber's Facebook post came from angry riders who demanded that the same be done to drivers-that their fares should be reduced if a driver turns up more than three minutes late.

Some also pointed out that Uber drivers tended to indicate that they have arrived at the pickup point prematurely, causing longer waiting times than necessary for riders.

It's easy to see how errant riders and drivers contributed to Uber's current unpleasant situation; it's all about psychology and selfishness, after all. Riders who request for an Uber before they're ready to leave are hoping to ensure that their driver is there the moment they reach their pickup point.

Conversely, drivers who indicate that they've arrived way too early are gaming the system to ensure that their riders are already at the pickup point when they eventually arrive. In both cases, the irresponsibility of humans resulted in at least one party's unpleasant waiting experience.

What's harder to spot, though, is that design has a part to play, too. You see, in an attempt to create an automagical car-riding experience, Uber's (and Grab's) design deliberately overpromises what it can do. And overpromises, as we know, lead to under-delivery and frustration.

To see how this happens, let's look at two crucial parts of the Uber/Grab car-riding experience:

• When the app tells you how far away your driver is
• When the app notifies you that your driver has arrived

Examining these key experience points would make clear the gaps between what the design communicates vs the reality of the app's underlying technology.

"YOUR DRIVER IS FIVE MINUTES AWAY!"

After you've booked a ride on Uber or Grab, the app gives you really useful information about when your driver will arrive.

WHAT THE DESIGN SAYS 

On Uber, a tooltip proclaims that your driver is "X minutes away." Grab offers similar information, giving you the time until your driver arrives. Note that in both apps, the design deliberately avoids any sense of uncertainty in the driver's time of arrival.

Ambiguous words like "estimated" and "approximately" are omitted and for good reason. In many situations, the sense of certainty improves the experience of hitching a ride.

THE REALITY 

The reality is, of course, that the estimated time of arrival is…well, estimated. The app uses the GPS on the driver's smartphone and mapping APIs to calculate the best estimate of the time it will take for the driver to reach the pickup point.

This estimate can go wrong for many reasons:

• The driver could make a wrong turn, extending the time it takes to reach his destination
• The driver could be under a large underpass or in a tunnel, cut off from GPS satellites. In this situation, his ETA will not be updated even if he's making progress.
• The driver could be completing a previous trip, resulting in additional detours
• The previous rider could have run out of cash (this happens more frequently than it should), and the driver has to either make a detour to an ATM or wait for the rider to withdraw cash
• Other unexpected and uncontrollable factors, like accidents, sudden heavy thunderstorms, software bugs, and dead batteries

[[nid:399007]]

DESIGN VS REALITY: ROUND 1

By eliminating the factors of uncertainty, the design is over-exaggerating the accuracy of the app's ETA predictions. In its attempt to calm the nerves of its riders, the ride-sharing app inadvertently sets up an overly high expectation.

Unfortunately, Murphy's law often prevails.

Imagine this scenario: You've just booked a ride on Uber/Grab and the app tells you the driver is seven minutes away. You've got quite a bit of time, which is great because you can get ready without having to rush. Three minutes pass, and suddenly your app pings you: your driver has arrived. Uh oh.

Or, picture this: You've managed to secure a ride, and the app tells you that you've got five minutes to get to the pickup point. You get ready and arrive at the pickup point on time. You think it's strange that the driver's not here yet. You open up the app and it says your driver is still four minutes away. What the…?

Sound familiar? (If your answer is "No," then you are one very lucky ride-sharer.)

Overpromising designs can drastically affect user expectations. When the design gets things wrong (i.e. an estimate turns out to be inaccurate), drivers or riders can appear to be errant even though they might be perfectly innocent.

In the case of truly irresponsible riders/drivers, overpromising designs can also amplify the implications of selfish behaviours.

"YOUR DRIVER HAS ARRIVED!"

Another key part of the Uber/Grab car-riding experience is when the app pushes a notification to tell you that your driver has arrived at the pickup point.

WHAT THE DESIGN SAYS

For both Uber and Grab, the design here is pretty simple. Once the driver has arrived at your pickup location, the app pings you automatically. If you're using Grab, the app even helpfully enlarges the car number so that it's easy for you to spot your ride.

THE REALITY

The app is much more manual than it seems. The technology to automagically detect when a driver has reached the pickup point is far from perfect. In fact, GPS can only be accurate to within 4.9m under an open sky, what more a highly urbanized environment like a city.

The solution? When your driver is close enough to the pickup point, the app allows them to manually ping the user whenever they've arrived.

Technological automation is deferred and manual human power is relied upon instead.

[[nid:399398]]

DESIGN VS REALITY: ROUND 2

This, of course, opens the system up to psychological gaming and abuse. Drivers are tempted to hit the "I'm here" button once it appears so they can ensure that their rider will be there the moment they arrive.

Compounding this problem is, of course, the problem of ETA inaccuracy discussed above. If the ETA on the app is inaccurate, riders might not be ready at the pickup point when the driver arrives.

If this happens frequently enough, it provides drivers more incentive to hit the "I'm here" button as soon as possible, even though they might not have any malice or intention of gaming the system.

Another compounding factor is the fact that geographical proximity is often not a good proxy for time-to-arrival, especially for cities. The car might be within a few meters of the pickup point, but complex road systems (or human errors) might result in a good three- to five-minute delay.

WHEN DESIGNS OVERPROMISE

By examining two key experience points of ride-sharing apps, we can see how designs that overpromise-with the intention to smoothen the user experience-tend to set themselves up for unrealistically high expectations and, eventually, frustration.

In the case of Uber and Grab, when technological inaccuracies and the inner workings of the app are hidden from view, what results is an (often incorrect) sense that it's the human that screwed things up.

When the app says the driver is there, but you don't see the vehicle, of course it's the driver's fault. And when a driver arrives at the pickup point to see no rider, of course it's because the rider is too entitled to get ready on time.

Most riders probably know that there is bound to be inaccuracies in arrival times. The problem is that most riders aren't constantly thinking about the technical inaccuracies of the GPS system.

In fact, the apps deliberately hope that riders (and drivers) will forget the uncertain, probabilistic nature of reality and believe in the soothing, simple message that your driver is definitely appearing at your doorstep in three minutes.

COULD WE GIVE HONESTY A TRY?

Is there a way out? The problem Uber and Grab face is a complex one: On the one hand, overpromising and oversimplifying things (as they currently do) will set up overly high expectations.

On the other, telling your customers that you don't actually control everything isn't very pleasant for both the company and its users.

But no matter how complex the problem, imposing penalties on users (whether drivers or riders) doesn't seem to solve it well. What it does instead is to further drive a wedge between drivers and riders, which can lead to the platforms' long-term instability.

Perhaps a viable solution could lie in having a more honest design. What if, once you've booked a ride, Grab shows this instead:

Honest design can mitigate unrealistic expectations and add a level of humility and empathy to the app. After all, on a fundamental level, most riders and drivers know that they're just people trying to make the best out of their day.

Drivers want to earn a living and riders want to get to their destinations; they both stand to gain by working together. As opposed to imposing additional fees, having an honest design does not automatically set drivers and riders in opposing camps.

Infusing honesty into designs can not only benefit the user flows highlighted in this article but also benefit other apps and companies, too. Sometimes, lowering the curtain to reveal the puppeteer can bring a nice human touch to the cool, smooth artificiality of apps or products that "just work."

Design has always been used to smoothen experiences, hide rough edges, and inadvertently overpromise what it can do. When it works, it appears magical. But when it doesn't (which tends to happen more frequently), it's a jarring return to reality that causes frustration.

Perhaps a move towards honest design will spark a new wave of more conscious, empathic, and compassionate people.

This article was first published on TechInAsia.com.