Uber and Postmates Shut Down Delivery Fulfillment for Restaurants
Why Consumer-Facing On-Demand Apps Can’t Provide a Reliable Logistics-as-a-Service Solution
“Isn’t Uber going to do this?” It was the same question every time. Almost every venture capital investor had concluded that all things logistics were going to be won by Uber. We were out raising money in 2015 for Homer Logistics (AKA Homer), and it wasn’t looking good. Homer is a tech-enabled services company that provides an outsourced delivery service for local businesses.
Uber had been exploding in valuation, and had recently started to show an appetite for things outside of ride sharing. They were painting a picture to their investors of using their massive fleet of independent contractors to not only transport people, but also to deliver everything from food to packages; and, Uber had started a service in mid 2014 called UberRUSH to do just that. Uber was targeting businesses and offering to deliver all of their goods to the end consumer, regardless of where the order had generated. UberRUSH made a huge push into major US markets with ad campaigns like this. My response back to the investor’s question was, “Uber is doing it, but it won’t work. Business-to-business (B2B)delivery fulfillment can’t be handled by a single-channel consumer-facing app.” Last week, an email went out to users that UberRUSH will no longer be open for restaurants. It’s shutting down.
A similar story played out for Postmates a year earlier. Postmates launched an API to allow businesses to “tap into their contractor fleet” and use them to fulfill their pickups and deliveries, even if the orders were not placed through the Postmates platform. Lots of consumer ordering channels ranging from Shopify to Dispatch used the API to help connect their orders to the contractor fleet. At first glance, this all makes sense… Postmates and Uber have massive fleets of independent contractors signed into their applications waiting to accept tasks. Here is why this idea failed.
Original Motivation: Stop Courier/Driver Churn
Uber, Postmates, Doordash, and other platforms have a massive challenge in their business model: turnover of couriers. Because these platforms have managed to classify their workforce as independent contractors, and not employees, they are able to pay people on a per-task basis. While this controls monetary losses for the platform when no orders are being placed, it also means that couriers don’t make any money when they are idle. At the end of the day, couriers are signed into the app in order to make money. This precarious balance of supply and demand is a costly endeavor, and if couriers aren’t busy, they quit signing into the app. As you might guess, this isn’t good for the platforms because if couriers aren’t available to run deliveries, the platform cannot reliably service consumer orders. As these bored couriers quite signing into the app, it also is expensive to keep hiring new couriers.
In order to solve this problem, Uber, Postmates and others tried to keep their couriers signed into the platform by providing “jobs”, even if the original order was not placed through the platform’s application.
The Unstable Nature of Demand Forecasting
While this seems like an elegant solution to stop churn in the courier supply, it wreaks havoc on the core business model. This is because tapping the same supply of couriers for both the platforms’ orders as well as the orders from other ordering platforms leads to terribly inconsistent service across all platforms.
In order to provide consistent delivery service to customers, and avoid courier turnover, there must be a very close match between supply of and demand for couriers. When a platform like Uber or Postmates limits its use of couriers to only service orders placed through their platform application, it has the best chance to solve this supply/demand problem. This is because Uber, knowing its courier supply, can best match these couriers to its order demands. However, when the platform is allowing the courier base to also service jobs from systems outside of the platform’s application, the task of matching supply and demand becomes much more difficult. The more the service demands from additional platforms vary, the more difficult it becomes to match courier supply and demand, and the more unreliable service becomes across all platforms, including the company’s core platform. More couriers on the platform does not solve the problem. This problem persists no matter how many couriers are signed into the platform.
Top Line Revenue and Marketing Fees
You might wonder why this matters. If the platform is making money by farming out their courier fleet to others, who cares? The explosion in food delivery has created a wave of companies running to grab consumer demand. These groups (Uber, Postmates, Amazon, Grubhub, etc.) all generate massive gross merchandise value (GMV) when they capture the full transaction value of the purchase, and charge the restaurant a “marketing” fee for sending them the order. These marketing fees now reach upwards of 35% of the order value, and are incredibly high-margin for the platforms. Investors love top line revenue, and a lot of companies are touting GMV as their top line revenue. It’s the main metric that the venture capital space looks at in terms of growth and traction. Even with questionable unit economics, venture capitalists have poured hundreds of millions of dollars into these businesses.
If you now glance at this supply/demand problem from the platform’s vantage point, the decision becomes very easy. Repeatability of delivery experience is suffering dramatically in the platform application because your couriers are being used by other systems. In Uber’s case, Uber ride sharing and UberEATS were using the same courier supply as UberRUSH. Postmates API was calling the same couriers as Postmates app. The GMV numbers and high-margin marketing fees are threatened because the customers using your application to order food are having terrible delivery experiences, in large part because your couriers are getting pulled into other jobs. These other jobs are low value transactions that don’t show investors attractive numbers and risk future financing and valuations. The image below shows the two scenarios and the total value back to the platform in each case. It’s an easy call: Shut it down. Don’t let other services use your courier supply.
A Void in Local Logistics
Platforms with a consumer facing channel cannot provide a repeatable, and acceptable delivery experience to local businesses if they try to utilize their courier fleets across multiple ordering channels. It’s a flawed approach and leads to instability in the platforms experience and jeopardizes their core business model. However, the need to provide local logistics solutions to businesses is still present. High volumes of delivery orders are being placed on channels without delivery fleets and sent to merchants at ever-increasing rates. There is an opportunity for a UPS or FedEx style operator to flourish in the local delivery space, and it won’t be solved by one of the large, consumer-facing applications you see today.