Lyft S-1 Form : How much do "incentive programs" for passengers/drivers actually cost?

Long-time readers of mine know that I’m not a big fan of accounting incentive programs as “sales & marketing expenses” (S&M) in platform startups.

The big assumption of accounting them in S&M is that they’re one-time for a certain new/returning user/client, that they’d generate long-term values or life-time values (LTV). In reality in certain platform businesses that lack inherent network effect such as ride-sharing apps, this is not the case. Passengers and drivers have no royalty and will just keep jumping around for the cheapest services, leading to service providers having to heavily subsidize the businesses.

The question is exactly how much do ride-sharing companies spend on such passenger/driver subsidy? And how much do the accounted S&M expenditures actually impact the unit economy?

In Lyft’s newly available S-1 Form, even though the company only qualitatively describe their accounting for such expenses and did not break down the actual numbers, they did announce the advertisement expenses out of the S&M accounts. If we assume that outside of advertisement expenses the S&M numbers are mostly incentive program expenses, then we can arrive at the following figures:

Inferring Lyft’s expenditures on incentive programs, a.k.a. subsidies

Inferring Lyft’s expenditures on incentive programs, a.k.a. subsidies

As can be seen, based on this inference, Lyft spent $451M on subsidies in 2018 to generate $2.157B of revenues. This is about 21% potential hit on the actual unit economy, if we agree with my argument that very little of the incentive programs generate meaningful LTVs.

Quarterly evolution of Contribution Margin of Lyft ( From S-1 Form )

Quarterly evolution of Contribution Margin of Lyft (From S-1 Form)

In S-1 Form, Lyft calculated its Contribution Margin as graphed above, improving toward 45.5% over the years. This 45.5% is quite close to its gross margin, which makes sense since a platform business has a very low fixed cost compared to variable costs.

If the unit economy is truly 45.5% then it’s not very bad. However, if we take my argument and further deduct the 21% incentives cost from the margin, we’re left with 24.5% actual Contribution Margin. Not exactly the sexy platform business that we all dream of.

One final point: the low-capital-efficiency nature of non-differentiating ride-sharing businesses also manifest itself in the extremely diluted ownership of the two co-founders of Lyft. Immediately prior to IPO, Logan Green and John Zimmer each holds only 0.49% of the company, compared to leading shareholders’ numbers such as Rakuten (13.05%), GM (7.76%), Fidelity (7.71%), A16Z (6.25%) and Alphabet (5.33%).

Still, it’s quite something that at least Lyft survived the capital battle and improved its business to where it is today. We all learn to be better founders and investors from their effort.

"We may never achieve profitability..."

Book value (contd.) - Warren Buffet said the same thing in his annual letter