Is Uber Assembling a ‘City OS’?
November 1, 2019 | Contributed by: Mike Boland
This is the latest in LSA’s Video Vault series. Running semi-weekly, it examines selected conference talks and video clips, including embedded video and key takeaways. Speakers’ opinions are their own. Check out the entire series here.
Uber’s success hinges on a certain network effect. Its two-sided marketplace grows as demand (riders) fuel supply (drivers) and vice versa. Gravitational pull towards supply/demand equilibrium keeps things humming along and works towards optimal supply-side liquidity (non-idle drivers).
Of course, Uber is still a great distance from profitability as it continues to reinvest in capturing market share in hundreds of local markets. That’s the name of the game in marketplace businesses. But the path to profitability is also about economies of scale through adjacent products.
This is the principle behind Uber Eats, where that market-share blitz engenders demand-side density that geographically clusters fulfillment in cost-efficient ways. That means bundling orders for adjacent services (rides, eats, grocery, etc.) to get variable revenue against fixed cost. That’s the definition of scalability.
With that macro backdrop, Uber is driven (excuse the pun) to a sort of “city OS,” which has interlocking, and economies-of-scale-boosting components. This naturally falls under the jurisdiction of “Uber Everything” director Jason Droege who unpacked the strategy at the Code conference (video below).
We’re working on grocery right now as a vertical. We think that’s complimentary with food delivery. And then you put rides in there and you sort of get a ‘city OS’ if you will, or a remote control for your city. And you can get all the things that you might transact within your market in… That’s going to be interesting, like how do we make that work? Where does that work? Does that work internationally? Does it work domestically? We have some ideas but we’re sort of forging ahead with that plan independent of where the stock price is.
The question of “where?” is logical as there are different dynamics in each market. That mostly goes back to supply/demand factors. An analogy here is how there’s a different tray of options in Uber rides (Pool, XL, Comfort, etc.) depending on which city you’re in. Uber Eats, Grocery and other products will be the same.
We’re in 36 countries and so there are business models that will work in Chile it might not work in the US. So we look market by market… For example, if you go to Costa Rica right now you can open up the Uber Eats app and there’s a toy store. The team just used the restaurant technology and they said ‘list your toys’ and some people use it. And so we’re constantly testing what works where. And it’s not a strategy: it’s just an experiment to inform the future.
Lastly, Droege touched on our new favorite topic: cloud kitchens (or what he calls “dark kitchens”). The delivery-only kitchens have cost and logistical advantages which we’ve examined. And Uber is leaning into the phenomenon for all those reasons, and because it’s a streamlined supply chain for food delivery.
A dark kitchen is basically a delivery-only kitchen. So it doesn’t have a storefront… Restaurants who get super popular can’t handle the in-store volume [so] they’ll actually pop-out and take cheaper real estate because they get so much volume. In that way, I think it’s fabulous but it’s up to the restaurant. We sometimes assist restaurant partners in finding space. But from a consumer experience, it’s just food from the storefront. So I think it’s just more nodes on the network if you will.
More interestingly, Uber is using its data to support these restaurant decisions. Given that it has so much consumption data from scaled operations, it can glean insights that help restaurants decide where to open cloud kitchens and how to dynamically alter the menu (as you can do in a cloud kitchen) to maximize yield.
We have this virtual restaurant program where we go to a restauranteur and say ‘hey here’s the cuisines or ideas in your market that aren’t being well-served.’ So a sushi shop might become a Poke bowl shop… Or a high-end Vietnamese place might offer lower [cost] Vietnamese under a different brand for delivery and so they can utilize that asset better than taking down a new lease. So I think there’s a number of models here. The important thing is that we’re getting consumers the food that they want and the actual production mechanism, we’re sort of neutral to.
See the full interview below. Also note that the above analysis deliberately sticks to economic aspects of Uber’s product and growth strategies and avoids other factors such as its company culture or political issues in play (though we recognize their importance and impact).