Uber’s Driver Problems Reveal Deeper Management Flaws

Something was just a bit off in the Wall Street Journal’s profile of Dara Khosrowshahi last week. The Uber CEO had been driving and delivering for the first time in his tenure and wanted to make the point on record. The Journal portrayed Khosrowshahi as a hands-on CEO learning the product with a splashy image of him in the driver’s seat. The largely positive framing was undermined, however, by the details in the story.

Khosrowshahi, the leader of a $63 billion company, admitted in the article he’d never used Uber as a driver until recently. And that, once he did, he found significant flaws. Fixing these issues was admirable, but waiting until nearly six years into his tenure to tackle them was unmissable. Attempting to convey progress, the profile left some analysts, investors and people close to the company wondering about its trajectory. 

“The WSJ article was a preposterous PR stunt,” Leonard Sherman, an executive in residence and adjunct professor at Columbia Business School, told me via email. “Dara is clearly on a driver charm offensive, trying to put lipstick on a pig.” 

While driving, Khosrowshahi found the driver signup experience frustrating and “clunky.” He learned that Uber wouldn’t clearly tell him when he was picking up two orders in one restaurant. He discovered that when he accepted a new delivery job, the app sent him to the new location even though he hadn’t completed his current order. And, as he drove, Khosrowshahi experienced a still-unsolved practice called tip-baiting, where customers write in big tips to expedite delivery and then reduce them once they have their food. 

An Uber spokesperson told me that the company had dedicated more resources to its driver app, and Khosrowshahi’s driving was part of an effort to improve. “As a company we were unfortunately more focused on the rider app than the driver app, until the post-pandemic driver shortage led to a reckoning and company-wide shift, which included increased on-the-ground attention from folks who didn’t already work on the driver app everyday,” the spokesperson said. “To illustrate that further, there are now four times as many Uber employees driving or delivering than there were before the pandemic.” 

Like every tech company, Uber is adjusting to the end of the Fed’s zero interest rate policy, and its focus on drivers is partially a result. When interest rates were near zero — and growth mattered more than profit margin — Uber could fix a subpar driving experience by luring drivers in with bonuses. In 2021, for instance, Uber announced a $250 million “stimulus” to get more drivers on the road. But now, due to market conditions, the company has less cash to throw around. So it has to focus on whether drivers actually enjoy working with it. 

Khosrowshahi’s appearance in the Wall Street Journal therefore seemed to address two audiences: One, Wall Street investors who worry Uber can’t attract drivers without cash bonuses, and two, drivers who might drive more via a better product experience. 

For some investors, the article landed with a thud. More than one pointed me to Khosrowshahi’s compensation, approximately $24 million last year, and wondered why Uber investors were 94% in favor of that package when the stock dropped approximately 40% last year. “There’s not strong alignment of interests with shareholders, which may cause their say-on-pay approval rate to come under pressure,” one Uber investor told me.

Khosrowshahi made 146.9% of his $2 million bonus for better-than-baseline company performance last year. After Salesforce and other tech firms felt pressure from activist investors, Uber might be another candidate.

As for the driver experience, many are fed up after years using a frustrating product. Some ex-employees suggested to me that Uber might’ve already churned through a good percentage of the population that would drive for the company. That Khosrowshahi himself had to drive to find some obvious flaws isn’t great either.

“What is going on in the product management part of this company?” Emil Michael, Uber’s former chief business officer, and no fan of Khosrowshahi, told me on the Big Technology Podcast this week. “If I were him, I’d go right to the product management team and some heads would probably roll.”

Uber has more to say on this, and I’ll print its response below. It also stands to benefit from Lyft’s struggles, where the share price is down 87% since its stock market debut and is moving forward without its founders. For now, I’ll end with this chart from Uber’s 10-K annual report, which plots the company’s performance since its 2019 IPO against the S&P 500. 

Uber’s response:

Since Uber was founded, we’ve encouraged everyone at the company to use as many of our products as possible to identify ways we can improve the user experience. That’s a key part of our culture that we’re proud of. Our Product and Ops teams — whose contributions are also discussed in that same WSJ story — lead that work day in and day out. But we think it should be part of everyone’s job, including the CEO, which is why we also have multiple company-wide Slack channels and employee beta testing dedicated to surfacing exactly this type of feedback.

Unlike most apps, our technology operates in the real world, which means product designs that can seem great when you’re at home or in an office can feel very different when you’re in traffic on a bike or behind the wheel with a passenger in the backseat. Our teams do thorough user testing with drivers to get their feedback during the design process. We also do validation testing of our products before we release them to make sure they are working as intended. But “dogfooding” our driver-facing products after they’re released is not as simple as debugging from a desk — you have to take the time to get a car, sign up, and get approved to drive or deliver. These real-world experiences can lead to additional discoveries, which then lead to additional improvements, and we think that is a key part of building the world’s best products in our category.

This isn’t some radically new process for us. But it’s something that we think has made a big impact more recently on the driver app, as discussed by the WSJ and this blog post from 2021. Even though things may be working as designed, we find many areas that can, indeed, be designed better. For example, as the article mentioned, we realized we could improve the interaction model when a courier receives a multiple-pickup order, or when they receive a new trip while they are still on a previous trip. We know there will always be things we can improve, and we’re committed to doing whatever it takes to deliver a great experience for everyone who uses our platform.