Delivery math, courtesy of  Uber Eats

Uber's CEO defends his company's role in the restaurant industry

Happy Inauguration Day!

Uber CEO Dara Khosrowshahi joined New York Times journalist Kara Swisher’s podcast last week. During the interview, he offered what seemed like a defense of his company’s practices and a rosy outlook for the future.  He says that the delivery business is now at a $40 billion run rate (run rates as metrics are so weird, but they’re one of the only ones that help make sense of a business’s trajectory in this strange, strange year. But they’re a projection — not an accomplishment!) Delivery bookings will be higher than rides in 2021, he says, but eventually he believes the business will settle in a 50-50 split.

He started the conversation explaining why references to the 30 percent fee Uber and others charge is wrong since that charge also includes the cost of what a restaurant would pay for the courier service: 

“The 30 percent doesn’t include the cost of the courier, right? And so if you’re looking at the economics that anyone is charging for delivery services, you’ve got to look at the cost of the courier and for example our Uber Eats service, our revenue take rate net of the courier cost this last quarter was 13%. And we think that’s a very, very reasonable take rate. At the same time, restaurants have a choice, if they want to use their own couriers, then the cost is typically 15 percent. If they want to use our service essentially for pickup, we charge them nothing and we offer them the ability to essentially build their own website. And if they’re bringing the customers, essentially to order, we charge them nothing. So when you look at that picture, 13 percent net of courier, bring your own courier 15 percent, pick up zero. Demand that you’re providing zero, we think that’s a very, very fair proposition and we are having restaurants sign up in record numbers. You could argue it’s because they have to.”

Swisher does argue that restaurants have to sign up because they’re left with little choice in the matter. Khosrowshahi again says he believes the economics are “more than fair.” 

Here’s the thing: the average consumer doesn’t care about the take rate, net of courier. The average consumer might not even know what a take rate is. The average consumer knows: 1. How much they paid for their food, and, 2. How many other charges they paid for the convenience of delivery in the form of delivery fees, gratuities, and other service charges. There’s a reason that these companies go straight to consumers when a city or other governing body attempts to intervene in their business model or pricing structure. 

This math also leaves restaurants out of the equation. Yes, it’s true, that if restaurants were maintaining their own delivery fleets (which they can do while still appearing on the Uber Eats platform), they would incur some delivery costs. Maybe they would be prohibitive. Maybe they’d have to pass higher fees onto customers instead. But they wouldn’t, in that case, be constrained by the bounds of a contract with a third-party behemoth. I’d argue they might charge a more appropriate, more realistic rate for the service they offer — one that isn’t offset by huge venture capital investments and valuation inflation on the public market. Of course, they’d also be missing a potentially large base of new customers. In this case, it’s up to the individual business to weigh pros and cons, but that math is even harder to do under pressure of failure in a global pandemic with spotty and unpredictable help from the United States government. (Here’s hoping that’s just one of the things that gets better after today.) 

Executives and investors, of course, do understand the importance of a take rate, and Uber has said publicly it’s working to increase the take rate on its path to profitability. (At the time, the target was 15 percent, so the company appears to be getting close. I assume they’ll then move the goalposts.) Uber probably won’t increase the amount it charges restaurants for the service, not outright. Too many localities have already moved to cap commissions restaurants pay during the pandemic, and some are considering extending them indefinitely in moves that are a real challenge to delivery’s business model. Khosrowshahi calls the fee caps “misguided.” According to a piece in the Wall Street Journal last week, regulators are “fed up” with the fees, and consumer protections could be on the table soon. 

Uber and others are more likely to work to increase the take rate based on things they can control: technology, efficiency, and product offerings. 

The problem is that “better” in tech terms doesn’t always translate to “better” in human capital terms. Take this recent Twitter thread from Vanessa Bain, an activist behind the Gig Worker Collective. It shows a message an Instacart worker received, and reads, “Our records indicate there are multiple instances of potentially fraudulent activity relating to the time taken to shop a batch or the distance driven to deliver a batch to your customer.” It goes on to encourage the contractor to “Do your best to efficiently shop and deliver batches. This keeps your customers happy and usually results in better ratings, too.” 


The full thread is worth reading for Bain’s thorough explanation on the topic and the so-called “algorithmic expectations” companies maintain. It’s just one example of the friction between technology and people, one that clearly shows that the tech platform values the customer’s satisfaction and feedback over it’s contractor’s experience. 

I’ve seen similar communication from Uber Eats to its partner restaurants, suggesting they not increase menu prices lest they risk losing business, referring to the company’s cache of data on what’s working well on the platform.

Khosrowshahi continues his defense of Uber Eats’ model with the familiar refrain: “We are bringing restaurants a whole new set of customers,” he says. This is largely true and has been since the beginning. Much of the business restaurants on Uber Eats and similar platforms receive is incremental, though as these companies become more popular, that might shift. 

DoorDash’s IPO filing underscored how important repeat business is for the service — the amount spent by repeat customers is growing at a faster rate than the amount spent by customers new to the platform. In the third quarter of last year, existing customers accounted for 85 percent of gross order value on the platform. Missing from that particular graph was any information about how many recurring DoorDash customers are using the service to order from the same restaurant repeatedly. 

That’s another part of the business I view as a potential weakness. A lot of restaurants can stomach a commission fee paid for reaching a new diner, that’s just the cost per acquisition. It’s a much harder sell when that diner becomes a repeat guest via a third party app.

There are companies working to address this, in what I truly believe is a sweet spot in restaurant tech right now. Abhinav Kapur, CEO of New York-based Bikky recently posted a Twitter thread about converting 600 guests of a local burger brand to direct ordering. According to the tweets, it resulted in $8,000 per month in incremental direct ordering revenue and a 17 percent increase in average check size. Perhaps more importantly, he says that half of guests place a second order after the initial conversion to direct ordering, and a quarter place a third order.  According to Kapur, 60 percent of third-party orders at restaurants Bikky works with come from repeat restaurant guests.

ChowNow, which offers restaurants a direct ordering platform and easy delivery integration for a flat monthly fee, has continued its push to encourage customers and restaurants that direct ordering is best, and its message is gaining traction among diners looking for an alternative to larger companies. Restaurants are also using online ordering tools like those from Toast and Bbot to accept orders directly, even if they remain on third-party platforms. 

Khosrowshahi claims Uber isn’t leaving restaurant owners in its dust, either. “We have pivoted more of our technical spend to actually help restaurant owners in terms of their IT and building out the functionality so that they independently can attract business, they can form a relationship with you, right. They can get their email, et cetera and they can reach out to customers directly,” he said. 

The full conversation is clearly worth a listen (or a read, if you’re me, team transcript 4ever). Khosrowshahi addresses gig work, intellectual property, the history of it all, and, of course, California’s Proposition 22. (He essentially refers to people who supported the initiative as those “with common sense” or a “normal person” though I encourage you to listen/read for context there.)

Khosrowshahi clearly believes deeply in his company and the future of on-demand delivery. In a business where speed seems to matter above all else, he wants Uber to own “next hour.” 

“And I think we’re in a good place now and the company has its own identity and we’ve got a bunch of people who are excited about building. Like if Amazon kind of owns next day right, with Prime, I think we can own next hour. You want to go anywhere in the next hour, you want anything delivered to you in the next hour, Uber can be the company that you can come to.” 

What else is happening? 

Deliveroo raised $180 million at a valuation of $7 billion. Amazon owns a minority stake in the UK-based company, which is expected to IPO later this year.  

Uber is looking to spin out Postmates’ autonomous delivery arm. TechCrunch reports that Postmates’ new parent company is seeking investors for its robot delivery business in hopes of turning it into a separate company. According to the piece, Uber will maintain a stake in the new company. It’s in line with the company’s previous decisions to sell off its autonomous vehicles and flying cars businesses. 

Kitchen United plans to open 16 new locations this year. The provider of space for commercial ghost kitchens will expand from its current four locations to new markets like New York and the Bay Area. 

Speaking of ghost kitchens, this well-reported piece from Eater Portland really nails the struggles that independent restaurants face as ghost kitchen operations scale. 

Slice, the delivery app for independent pizzerias, introduced a “pizza score” for its restaurant customers as a way to measure their success. Restaurants rank in one of four categories from “okay” to “all star.” It’s a restaurant-friendly tool, especially for business who may be new to the online ordering market. The simpler the better, I think. 

Pizza Hut Israel is working in partnership with a tech company to deliver via drone. The drones would expand delivery radiuses by moving food from restaurant to courier, not dropping it directly with customers. Dragontail Systems, the technology behind the drones, says it uses AI technology to “optimize and manage the entire food preparation process form order to delivery.” It’s those algorithms again.