The car-as-a-service vs e-bike-as-a-service model breakdown

Here’s some data I found in the Swedish media detailing the car-as-a-service model breakdown for Volvo’s M service:- period: 2018 - mid 2020 (30 months). The numbers are aggregated.- 70k registered users (nota bene: not payers!) - 5 cities: Stockholm, Gothenburg, Malmö, Lund and Uppsala- investments of SEK 640M ($72M) - losses of 400M ($45M) - revenue 242M ($27M)

In case you didn’t know, M is an on demand car service operating on a hybrid model: roundtrip, station-based, with a subscription fee and pay-as-you-go-expenses on top.

The user number is from a longer period actually, since M was launched in 2019 as a rebrand of and replacing an initial service called Sunfleet, which was more 30% more affordable - the transition likely generated churn.

The obvious difference of car-as-a-service compared to ebike-as-a-service is that instead of having to optimize for an ARPU spread over a $30-40k asset during a 3-4 years lifecycle, you optimize for a $2000 asset at a 2-3 years lifecycle.

Quick back on the envelope:

- at a monthly ARPU of $500 we get almost 5000 customers. Since the revenue is cumulated, the number is lower than that.
- a fleet of 1000 cars @ 30k accounting value at a monthly ARPU of $500 is covered by 5000 users, before SGA expenses.
- parking spaces are also big items on the income statement.

The obvious difference of car-as-a-service compared to ebike-as-a-service is that instead of having to optimize for an ARPU spread over a $30k asset during a 3-4 years lifecycle, you optimize for a $2000 asset at a 2-3 years lifecycle. 

And so, for a fleet of 1000 e-bikes at $2k a piece at a monthly ARPU of $59 you need less than 3000 users to cover inventory. 

And in a sub model you don’t have to budget parking/docking.

Here is a detailed analysis about the car as a service breakdown in the Nordics I wrote some time ago.

No comments yet

CAPTCHA