Amazon Flex Could Be The Next AWS

| About:, Inc. (AMZN)


Rideshare networks can be very profitable once completed, but often cost billions to start up.

Amazon's existing package delivery network might allow it to save on many of those start-up costs.

The potential profits of such a mature delivery/taxi network could be very large.


Amazon (NASDAQ:AMZN) CEO Jeff Bezos has famously said that Amazon has three great businesses and is looking for a fourth. I think they may have already found it. No, I am not talking about their Echo home speaker/computer/assistant, though that does seem to be doing well, as does their Fire tablet platform. I am talking about their new logistics investment. No, not the airplanes. The other one.

Seeds Of The Future

Prime Now by all accounts has been an expensive drain on the company finances, but Bezos is famous for taking the long view and funding ventures for years with fresh capital waiting for it to grow into something. That is how Amazon, a book-selling retailer, became the world leader in cloud computing. And they may soon be the leader in another seemingly unrelated category. Because to help fulfill Prime Now's promise of as little as one-hour delivery, Amazon also launched Flex.

Amazon Flex is essentially the equivalent of Uber (Private:UBER) or Lyft (Private:LYFT) for package delivery. Amazon hires drivers who use their own vehicles to deliver Now packages in select cities. Launching first in Seattle, the retailer's home base, the service has since expanded to many major markets including Chicago, New York, multiple Texas urban centers, and more.

Many have already remarked that Flex offers Amazon the opportunity to, in the company's own words, " change the way Amazon moves and delivers packages." One of Amazon's biggest cost components is shipping charges, and a lot of those are incurred at the "last-mile" of delivery. If directly hiring drivers in their own cars can help bring that down, then by all means. But I believe this is a case of missing the forest for the trees. Amazon can do much more than move packages with this platform. It can move people, and potentially do so more efficiently than Uber or Lyft can.

Amazon's Inside Track

Uber and Lyft may be very profitable once they get up and running, but they have to burn hundreds of millions to billions of dollars getting themselves up and running. You can read my detailed explanation as to why, but the short version is that the nature of the "network effect" means that they have to guarantee driver pay when they first set up the network, often far higher than what they can hope to recoup at first. They need to offer fast, convenient pickup with short wait times, or they will never get passengers. And that means that they cannot wait until they have passengers to start paying drivers. Once the network is up and running the passengers are their own guarantee, basically, and the tech companies can start raking in the commissions.

These high start-up costs are why so many Uber competitors have failed, and why most of those now thinking of entering rideshare are established companies with large cash flows who can absorb the hit of the startup costs, like Ford (NYSE:F) and GM (NYSE:GM).

But Amazon may be able to skip this very expensive trip up the hill, and get the smooth glide down the other side essentially for free. With Prime Now on the books and guaranteed two-hour delivery already promised to tens of millions of Prime members, Amazon will soon have probably thousands to tens of thousands of cars and SUVs on the road delivering Now packages.

The details of Amazon's delivery algorithms and fulfillment procedures are proprietary, of course. But its not hard to make an educated guess of the broad outlines. Presumably as Prime Now orders come in, Amazon tallies all the orders in a given part of the city and designs the most efficient route a vehicle could take to deliver packages to a series of customers.

When the cars are done making their deliveries, they have to return to the SC to pick up another group of packages. For a hyper-efficient system like Amazon, this is the most maddening part. Those are empty cars, burning time and gas while they travel back from wherever they ended up to the SC for another load, making no money while they do.

Amazon has empty cars with no business, and as more and more people come to socially accept ridesharing services, there are more and more people looking for empty seats in cars going their way they can hitch a ride on. My guess is Jeff Bezos has already put two and two together. By giving out rides in empty vehicles that have finished their deliveries, he can reduce downtime and monetize what had been a cash hole in the business model.

Filtering Requests

Obviously, it would be inconvenient to have a rider who wanted to take the car in the opposite direction away from the SC, where more packages were waiting. But that's an easy problem to solve. Before it shutdown, Sidecar, another ridesharing service, allowed its drivers to accept or reject rides based on destination. They could go about their own business and only pick up people who were going their direction. Lyft now has something similar, which it calls Destination Filter. Uber does not yet, but I would assume its an easy tweak if it proved popular.

Amazon really doesn't care if its popular, though it probably would be. For Amazon, what's important is that Destination Filter can be very efficient. Even more so for them than for Lyft. By screening ride requests for empty cars to only look for passengers headed towards the SC, it can keep the detours and added trip time to a minimum while still monetizing the empty seats in the car.

Of course, that kind of limit would mean that Flex was only offering rides to a fraction of the people in a city, the ones whose destination was in the direction of the SC. But the app does not have to limit itself to offering rides in empty cars.

Plus The Other Half

As I said, all networks have start-up costs, and that is as true for Prime Now as it is for Uber. Ideally, Amazon would have multiple Prime Now package deliveries clustered as close together as possible, allowing it to maximize the amount of packages each car delivers in a two-hour window and minimize the cost per delivery. But since Amazon is guaranteeing two-hour delivery, it has to send out cars by a certain time, even if they do not yet have enough orders to make the run economical.

A ridesharing platform alongside the shipping platform could enable Amazon to convert these "half-paying" car trips into full-service ones by integrating ridesharing not just into the return leg of the trip, but into the delivery route itself. If a car leaves with half a load of packages, it presumably has some extra time to pick up a passenger and still make its two-hour window for all packages. Amazon could once again use destination filtering to make sure the passenger's route did not take it too far out of the route it had mapped out for deliveries.

Obviously, since the packages are still in the car, Flex might have to be a little less generous in terms of space for passengers. The Flex app might warn a passenger that the number of riders in their party was being capped at three instead of four, since the packages are in the passenger seat, or that only half the trunk was available for their bags on that trip to the airport. This would be an inconvenience for a small fraction of the potential riders, but most do not need four seats or the whole trunk and probably would not mind. Anyway, Flex would presumably be compensating customers for the inconvenience by offering savings on price, since they have less downtime than Uber drivers and can charge less per-minute and per-mile.

Even with those customer savings, the potential profits are very large. Uber is already sporting a $62.5 billion valuation. If Amazon can do the job more efficiently than Uber its own Flex might be worth even more. Amazon Web Services' (AWS) value was projected at $50 billion two years ago. Even if it has gone up since then, Flex and AWS would at least belong in the same conversation.


Uber may be the next $100 billion tech titan, but it is also a very capital intensive one, especially for a business that does not actually own any vehicles. Uber is basically just an app, plus the accompanying customer support. And yet thanks to start-up costs it's burned through billions trying to get itself off the ground. A platform which allowed Amazon to save most or all of the driver incentive costs, and save on start-up costs for its own new package delivery service as well, could give Amazon a powerful - and sustainable - competitive advantage. Although Ford and GM are talking louder about mobility, it may be Amazon that becomes the third major player in the P2P transportation market.

Disclosure: I/we have no positions in any stocks mentioned, and no plans to initiate any positions within the next 72 hours.

I wrote this article myself, and it expresses my own opinions. I am not receiving compensation for it (other than from Seeking Alpha). I have no business relationship with any company whose stock is mentioned in this article.

About this article:

Author payment: $35 + $0.01/page view. Authors of PRO articles receive a minimum guaranteed payment of $150-500. Become a contributor »
Problem with this article? Please tell us. Disagree with this article? .