Uber, today's hottest tech startup, has reached gargantuan status on the premise of getting a ride ‘at the tap of a button’. With over 2 billion trips taken as of July 2016, they’ve very well continued to make good on their promise with improvements and experimentations. It’s crazy to think in the not too distant future, Uber might operate a fleet of self-driving cars that can run 24 hours and arrive exactly when you need it. However, we’re not there yet. Today, across 400+ cities that Uber is active in, over a million daily rides are fulfilled by a network of a million (human) drivers, including me. As a broke college student looking to make extra cash, I started driving for Uber part-time, and it’s been clear to see that Uber’s operations take place not in city street and highways, but in the private cars and smartphones of its driver network. It’s baffling, then, to wonder why I've spent the last 6 months cursing about how much the Uber Driver app can be improved to offer a better experience to its users.
My application for the KPCB design fellowship challenged me to pick a company in their portfolio and redesign a feature of their product. So, with the unique position of being a product designer with a stellar Uber driver rating, I leveraged the Nir Eyal Hook Model and Information Foraging Theory to redesign three features of the app. By growing a happier and more predictable driving fleet, these features are in line with Uber’s business objectives. Here, I go into detail about the experience of being an Uber driver, drill into some of the problems that drivers endure, and examine how Uber's business models affect the driver experience. I then explain how understanding context and giving peace of mind can be powerful agents to help drivers feel like first class citizens on the Uber platform.
Disclaimer 1: Interaction and visual design do not exist in a vacuum. In this redesign, I attempt to round out the whole user experience by working in line with the limits of technology and my limited knowledge of Uber’s business objectives.
Disclaimer 2: It’s important to note that the redesign of this app was based on my experience using the app for the past 6 months. To be able to have found other Uber drivers to get feedback from, I would have had to pay for Uber rides that I didn’t need.
How's life as an Uber Driver? I'll tell you!
First and foremost, it’s important to understand the fact that drivers are in-fact operating a motor vehicle in what’s likely to be high traffic hours. Below is a video of one of my favorite comedian, Louis CK, eloquently describing his thought process in traffic. Please entertain yourself with it. (Viewer discretion is advised).
Uber Drivers go through quite an emotional rollercoaster on their driving sessions; enduring multiple emotional states between dealing with traffic, weather conditions, demand fluctuations, driving alone and picking up a rider (read: stranger) who may or may not want to engage in a conversation.
Per the current Uber model, drivers aren’t paid for the distance they drive to pick up a rider or the time it takes them to get there. In the same vain, there’s definitely a cognitive dissonance associated with drivers getting to a pickup location before the rider is out or even ready. After all, time is money, and time spent wasted waiting for a rider is potentially money lost not fulfilling another trip.
Funny enough, some of the biggest headaches drivers have, comes from the very interface they use to conduct their business. The Uber Driver app has a few problems itself, but there’s also the ecosystem of apps that drivers need to chart their destination, communicate with riders en-route, and entertain them once picked up. While switching between these apps, drivers’ phones can often slow down, lock up, and even run low on battery. This can be a terrifying experience for a driver who’s tasked with maintaining the highest rating possible in the middle of a multi-stage transaction that that might involve: driving in the rain to an unfamiliar location, letting a drunk stranger into their privately owned vehicle, and battling traffic towards their destination - the only burrito joint in the area open past 12. Lovely.
When a business tactic shoots itself in the foot
Uber, like any startup, has to optimize for business growth objectives. For Uber, this mainly boils down to getting as many drivers on the network as possible and keeping those drivers behind the wheel long enough to match demand. As someone who’s fulfilled well over 100 Uber trips, I’d argue that some of Uber’s business objectives have found their way down to product design tactics that actually work against keeping drivers driving. They certainly don't alleviate the problems intrinsic to this kind of vocation. One example is the flurry of text messages from the community managers who try to coax drivers into getting online during high demand times. Another example is forcing drivers to accept trips while being kept blind to the drop-off destination, only being revealed after they’ve picked up the rider.
I can tell you, after a 2 hour session, I might go offline simply because I'd rather not risk the possibility of accepting another ride that could take me 45 minutes away from the next thing on my personal schedule.
Imagine for a second, a future when Uber manages a self-driving fleet of cars. There are many ways autonomous vehicles would improve the service. But what if each vehicle in this fleet went online and offline at random intervals, this erratic fleet behavior would largely diminish the logistical benefits of an autonomous fleet. My point is that you can bet your dollar that Uber would pool their resources to understand why exactly the active time of their fleet couldn’t be controlled or predicted.
In the same vein, its important to acknowledge that the accumulation of all the above mentioned frustrations certainly adds up over the course of a driver's session, and the logistical effect is similar to the analogy I just mentioned. The accumulation results in the fatigue of a driver's willingness to prolong their driving session, to ‘get back in’ more often, or even continue doing Uber at all. The lack of understanding of what drivers' endure is a weakness in Uber’s business model and threatens their profitability.
The big concern here is the context - personal goals and emotional states - of the driver network. These shouldn't be ignored so much as oil changes and low tire pressure could be ignored in a self-driving fleet. Uber’s technology platform has been nothing short of innovative in handling logistics and reliability, but there’s another way Uber can claim innovation. Uber can set an example for how gig-economy startups - with over $120 million in VC funding in 2016 alone - value the experiences had by their supply of workers.
With this in mind, I propose three features that can improve Uber’s models - and profitability - by understanding the factors that prompt a driver to go online, keep them in their session, and cause them to go offline. By understanding their context and giving them peace of mind, Calendar Sync, Online Modes, and the new Rider Card can enable drivers to earn more and do so more comfortably.
Getting them invested: Calendar Sync
By keeping their emotional state in mind, we can take action to improve their driving experience which might keep them driving longer and more often. I tapped into some behavioral models that might get drivers going online more often. Specifically, the Nir Eyal Hook Model that some of the most habit-forming apps like Pokemon GO have replicated with great success. The Driver app certainly accounts for Action - when a driver opts to go online - as well as Reward, as they earn money on sessions. However, it doesn’t really allow for Investment or use any effective Trigger that could develop habits.
Admittedly, without concrete data to back this up, I would guess that most people who drive Uber, do it around their personal lives. They aren't Uber driving machines who happen to have personal matters; at least they shouldn't be seen that way. Calendar Sync lets drivers get more invested when they import their schedule onto the platform. This lets the app get a read on the times they might be available to go online and for how long. Now that a driver is invested with anticipation of their next available driving session, you can later ‘trigger’ them with a notification 15 minutes before the next gap in their schedule.
Understanding Context: Online Modes
I took the current 'Go Online' UI element and fleshed it out into a new feature that allows a driver to specify which mode they would like to begin a session in: Destination Bound, Time Limited, or Earnings Motivated.
In Destination mode, drivers will only be given trips towards destinations that are along their specified path. The current Uber app does allow drivers to do this, but it's a separate UI element hidden away from the regular Online button. I designed the go online feature to allow drivers to search and save more locations than just ‘home’, and even recommend locations based on Calendar Sync.
Over time, the flexibility of these online modes makes drivers associate being in their car with making money, thus going online more often and in more situations than they currently think to. Even though as a company, this is very limiting for assigning trips to these drivers, this is ultimately a net gain of drivers on the platform at any given time. With more data from drivers about their motivations, Uber can better improve their models on how likely the company is to capitalize on demand and adjust their messaging on recruiting new drivers.
Giving Peace of Mind: The Rider Card
As an Uber driver, When I get to a pick-up location and the rider isn’t outside ready to be picked up, my first instinct (after a big sigh!) is to call them to get an idea of their status. Unfortunately, the current navigation hides the rider’s number and requires at least 5 taps to get to the point of dialing the rider's number with the phone app. This is an unacceptable level of friction for any consumer mobile app, not to talk of one that’s used while operating a motor vehicle.
Regarding this same scenario, the current app does do something helpful. When I arrive at the pickup location, it lets me know that a notification has been sent to the rider. That’s great for the rider, and it is nice for me to see, but it doesn’t give me much information regarding how long I have to wait in this ‘No Pickup. No Drop Off’ zone that surrounds this downtown hotel like a moat. This is all too baffling considering that there already is a feature that is primed to solved this problem, but it falls far short of doing so. The current app shows me a rider card when I'm in the middle of a trip. But this card shows only one thing, the name of the rider. That’s all folks!
Information Foraging Theory is the concept that relates information navigability to the cost benefit analysis that an animal does to weigh the energy to catch a prey against the energy it will gain. The idea here is to account for information requirements and to effectively group content where the user can best find it.
So I designed the Rider card to include not only that the notification has been sent to the rider, but also to show whether it has been acknowledged as well as an update on the rider’s status that they are either: ‘Outside Waiting’, ‘On the Way’, or ‘Running Behind’. This is very valuable information to a driver that might only require a couple taps on the rider’s end. But wait, there's more! I went even further with the card design, and made it easy to call or message the rider. With this redesigned card, a driver may never actually need to call a Rider, but if they ever needed to, they could initiate a call ‘at the tap of a button’.
I enjoy asking questions and drilling into problems, so to gain more feedback on the current Uber app and better inform my redesign, I really wish I had 30 to 40 full time and part time Uber drivers at my disposal for interviews. At the end of the day, designing in the dark - without user feedback - is just betting on your own perspective.
Reducing friction is one of the core values of human-centered design, and this redesign further impressed that upon me by challenging me to keep in mind all of the stakeholders related to a product ecosystem. We're all humans with limits and capabilities, so as designers its important to account for that.
This redesign also taught me the challenges of designing within the context of business objectives. As a cross-disciplinary student, I hope to gain more experience in balancing business goals and technological capabilities against the prime goal of user-centered design.