UBER masks this real number across many of their general line items.

EG they put ~$300,000,000 of the "Driver Incentive" cost into "Cost of Revenue" line item that represents:

- "Any amount paid to a driver that exceeds the revenue earned by that driver (for instance, if a driver’s earnings from a trip exceed the fare for that trip). Excess driver incentives jumped by about $300 million in 2018 from the previous year, largely due to Uber Eats."

And then they stick $1,800,000,000 into "Sales & Marketing" representing:

- "Discounts, promotions, refunds, and credits for customers,"

- - - -

A popular criticism of UBER is that their unit economics are not sound - eg they can't make money at scale on a per ride basis.

So this giant $2.1b glut of Driver Incentives and Customer Incentives is one of the areas to deeply study and build a position around if you'd like to be an UBER stock holder :)

In terms of unit economics, UberX is still (mostly) sound -- the rider is paying a fare based roughly on time/distance, and the driver is paid a fare that's also based roughly on time/distance; it's very rare to see a rider paying less than what a driver gets paid. (Even after taking into account the various driver incentives.)

Where the unit economics get completely messed up is Uber Eats. In my local market (Wellington NZ) the eaters are paying a flat delivery fee of $5.99 to $7.99, while the drivers are being paid $5 per pickup and $1.35 per km. This results in drivers earning anywhere from $5 to $20 for each delivery -- and that's without even taking into account any driver promotions. No wonder Uber had to come up with an "excess driver incentives" line item!

Sometimes Eats will have the driver pick up two orders at the same time. This lets Uber collect two delivery fees from the customers while only paying the driver a single pickup fee plus the kms from restaurant to customer A to customer B. In theory this should work in Uber's favour. But their system all-too-frequently will make insane assignments, such as batching together two orders for customers that live in completely opposite directions. There will be situations where the customers have each paid, say, $7.99 for delivery... and the driver ends up earning $30 from the trip!

So this is cutting very deeply into the 35% commission that Uber charges restaurants on the total order value. They're trying to hide the insane unit economics of Eats by spreading the various revenues and costs around different line items...

Back the day I had a short stint at a bus charter business. There we also tried to reduce costs by bundling travel group's individual itenaries and match that against available busses to reduce empty rides. The result would have been that groups might have different busses on different days. Nice idea, in theory.

Turned out that avoiding empty trips in transportation is hugely dependant on volume by time and region. It is also a very hard problem to solve as you need to match availability data from busses (which needs to be collected in a scaleable way), trips (which need to be in the correct regions and times) and then simply come up with an optimization algorithm to sort it all out. Peace of cake... Not. Especially if you want to link that to your pricing already during customer booking instead of after-the-fact discounts.

One might suspect that the transportation industry which has the volumes and the cost pain of empty trips might already have solved it if it woupd have been that easy.

The elephant in the room for ride sharing is unit economics (and, closely related to unit economics, the race to self-driving).

Hiding $1800m of ride discounts under Sales and Marketing feels deceptive. If a store advertises "summer sale, 30% off!" for a month, you can reasonably call that marketing spend.

Uber is different. If my flight's leaving in 1h30m, I pull out my phone, open Uber, type in LAX, open Lyft, type in LAX.

Say Uber is $17. Lyft is $20. I pick Uber. If there's a "promotion" going on under the hood to produce that $17 number, is that marketing? It's really just price competition in a cutthroat, negative-unit-economic market.

Given that, Uber laying off 400 people has a dual effect. It's cutting sales spend. It's also creating news to distract from the fact that >50% of the sales spend is really per-ride losses.

Really good point there. If Uber can charge $1.01 for something it pays $0.99 for, then it might have a sound business model. But if it's in a perpetual state of offering discounts down to $0.98 to keep users buying rides, then it probably doesn't, and to relabel the underpricing as a marketing effort is just obscuring that.

I am not a shareholder, but I think market share is the short term goal. The incentives can be phased out as prices increase a bit after competition cools off.

A criticism of this "model" is that building a ridesharing app and cloud backend is now not that hard.

So some developers with VC money can develop an app, launch in a city, and force Uber to cut prices, because people will be happy to install a FastRyde app and get rides for $3 less than Lyft/Uber.

And that is exactly what is happening in Paris, for instance. There are a ton of different apps there. Kapten is one of them. They simply cut their commission by 15% compared to Uber, pay the drivers 7.5% more and make the rides 7.5% cheaper, basically. Since they don't have as much legal costs and marketing costs as Uber, it seems to work out for now. I really don't see how Uber will be able to prevent this in other places too...

The big question is driver retention and how much Uber subsidizing drivers is required. Maybe Uber spends 0 extra, drivers drop out of the system, ride prices increase, and drivers return as riders stop using Uber, bringing it to some sort of equilibrium. Maybe not.

These retention numbers, for any business that needs returning customers or service providers pretty much make or break all the customer/provider acquisition economics.

In London Uber is unravelling already. Here Citymapper is the default transit app and Uber is just one of the options listed. The transfer cost for a customer changing app is 0. We keep saying we’re “not sure” whether Uber will have customer retention... no, of course it won’t! Unless they change their model, they’re the biggest VC folly in recent times. Just proves how overheated the market become when F/A/G acquired everything in the market that competes with them, leaving ridiculous Vc money to flow to dumb ideas like Uber.

no VC money anylonger, now it's stock market money, i.e. our pension funds

Wait, you guys get pension funds?

> The big question is driver retention and how much Uber subsidizing drivers is required.

In Australia we've got drivers advertising other services directly to riders. Most of these are probably subsidizing rides to enter the market, but how long can uber continue losing money as competitors try and enter the market?

> market share is the short term goal.

Creating a monopoly then jacking up the price. Not doubt shafting drivers and their customers.

But Uber's business model doesn't have lock-in potential. So if a price change happens, people (drivers and customers) will just switch to a different app.

I have sympathy with your view. Do Uber even care about that? It's selling the prospect of a lock-in that got them the IPO.

To me this only works out if they are a monopoly bribing politicians to allow them to keep being the only smartphone-app-based taxi and food delivery service in town.

As others mentioned, its not hard to clone Uber. They don't even have the network effect of Facebook or Twitter. No one cares if their friend uses Lyft and they use Uber, they just care about the one time they make a journey, and how much it costs (along with safety I suppose, but that is an extra cost Uber try very hard to avoid).

The drivers are free to run whatever apps they want simultaneously, because Uber only works when exploiting them as "contractors" free to choose when and how they work.

I think Travis Kalanick was right, they need self driving cars, now...

From their S-1:

>Driver incentives. Driver incentives refer to payments that we make to Drivers, which are separate from and in addition to the Driver’s portion of the fare paid by the consumer. For example, Driver incentives could include payments we make to Drivers should they choose to take advantage of an incentive offer and complete a consecutive number of trips or a cumulative number of trips on the platform over a defined period of time. Driver incentives are recorded as a reduction of revenue to the extent they are not excess Driver incentives (as defined below).

>Driver referrals. Driver referrals refer to payments that we make to existing Drivers to refer new Drivers onto our platform. Driver referrals are recorded in sales and marketing expenses, as they represent the receipt of a distinct service of customer acquisition for which there is evidence of fair value.

https://www.sec.gov/Archives/edgar/data/1543151/000119312519... (pp. iii)