Scooters! — VietEcon

Ok, got to be honest here, this is not a post about Vietnam. I am endlessly fascinated by the sharing economy, and especially the companies that are in public transportation. In Vietnam, that’s mainly Grab and Go-JEK. Grab has the pole position, but Go-JEK just hired the former head of Facebook Vietnam (who co-founded Misfit, acquired by Fossil in 2015). Plus, there are a lot of smaller companies that are competing. These all have interesting business models that may not be sustainable. I think that the cost of switching these ride-hailing services (from the passenger and the driver perspective) is so low that there is really no barrier to entry. That means if a player pushes on prices or wages, they will win, at least in the short term.

But so far, none of these companies have brought in scooters to Vietnam, which have been the big new thing in the US and Europe. Bird went from nothing to 10 million rides in one year. None of the companies in Vietnam have these scooters yet, but I could see this as a potential market over time, if the unit economics work.

Oh, and scooters do not mean mopeds or bikes, but the tiny things that used to be the exclusive domain of children.

The big stumbling block is the economics on a per unit basis. According to some research, scooters at the beginning of the trend had revenues of just $75 over their lifetime, but cost $550. That is not good. That’s a big $475 loss per scooter.

To answer this concern, a VC, Mark Suster, wrote this Medium post. He is an investor in Bird, the largest scooter company in the US, so he is definitely “talking his book.” His view is that unit economics have been bad, but they are going to get much better. Plus, the technology will be so good that it will mean both a profit and barriers to entry for new competitors. Plus, while it is capital intensive at first, over time that will not be true. A few points:

I feel like everyone at the beginning of this journey thought that scooters would be like other ride-hailing companies. Not capital intensive, but then realized they need to provide all the scooters. I am sure that is not true, but it kind of seems like it. They need to buy lots of scooters, then buy lots more scooters when those fall apart.
In the US, there are about 250 million people that live in urban areas, but if you look at those that live in dense neighborhoods, it’s more like 20% of this, or 50 million. If 5% of these people take a ride once a day (which seems fairly high), that would be 2.5 million riders a day.
Assuming each rider averages 2 rides per day, that would mean 5 million rides. Scooters generally make 5 rides a day now, according to data from the Information and Louisville. That means there need to be 1 million scooters on the road at any time. Over time this will likely fall, as the average utilization could rise from 5 rides per day to 10.
Scooters are currently expensive. Bird said that it spent $550 at the beginning, and that it wants that to fall to $360 for a more durable scooter. Over time, these costs may fall (as battery prices fall) or increase in order to make them more durable. Suster doesn’t really talk about scooter costs, just that the new scooters are 60% more efficient and durable.
If there need to be 1 million scooters at an average cost of $360, then that’s $360m for scooters at a time. Assuming these last 3 months, the companies need to spend $1.4bn per year. This could fall to $720m as rides per scooter per day increases and then down to $360m as the life of the scooter increases to 6 months.

We could easily reach a situation where gross margins for scooters are high (based on certain assumptions, I could see them reach 32% after depreciation. And if the numbers are large enough, that would cover staff and R&D (operating expenses outside of COGS). See table above (as you move left, I make changes to the costs/revenues to reflect improvements to see what has the largest impact - at the end, we get high gross profit per scooter).

But one of the problems is that we are going to have some people that ride a lot. Say 3-4 times a day, 5 times a week. If the cost to that person is the current revenue of $3.65 to the scooter company, that means the weekly cost is $73. You can buy a scooter for just 7 weeks of riding. So your power users could just buy their own to save money. Maybe this would only be a small problem, but it would mean that some of your best users may leave the service.

Otherwise, completely outsourcing the manufacturing of scooters, like companies did at the beginning, is not a good model. Bird has to invest, with its manufacturing partner, lots of time and effort to get the right scooter. Other companies will have to do the same, and that will be expensive. Bird is also looking to franchise the model into other countries, in order to lower their capex. Given current unit economics, paying another 20% of revenue would be unprofitable.

Ultimately, I thought that Suster put out a compelling narrative but without numbers. I think it will be interesting how all of this changes over time. I have ridden a scooter in Santa Monica (highly recommend) but it was extremely expensive (more expensive than for an Uber ride over the same ground). I think there is a place for scooters, and it will probably be in cities with good weather, public transport that is good but not amazing (so one needs a scooter for the last mile), and at a reasonable price.