4 ms·
A 20-30% gross margin is incredibly thin. It's smaller than brick and mortar retail, which is widely considered one of the most competitive industries. The vast
by bsbechtel 11y ago
A 20-30% gross margin is incredibly thin. It's smaller than brick and mortar retail, which is widely considered one of the most competitive industries. The vast majority of companies that operate on margins like this have low $/sq ft retail or warehouse space that they operate out of, pay employees at or below market salaries, and do many other things to keep overhead costs low. I don't know any details on Instacart's operations, they could be incredibly thrifty in how they spend money.
However, the vast majority of startups operating in the on-demand space have offices in downtown SF, one of the most expensive/sq ft places in the country to open an office. Additionally, engineer salaries in SF are averaging >$150,000/yr. Contract this with Amazon, a retail giant (<35% gm) who has yet to turn a real profit and is operating at a scale most startups can only dream of achieving. Amazon based their operations in significantly cheaper Seattle, and Bezos started the company giving every employee a desk make out of old doors and 2"x4"s to instill a culture of thriftiness into the core of the company. Sorry, but I just don't see how many of these on-demand startups can really make it when they are spending money at the rate they are with the cut (margin) they are taking.
- Naritai 11y agoTo add to your analysis, Instacart is also very dependent on a large team of 'staff' (be it contractors or employees) to provide its service. A major cornerstone of low-margin companies like Amazon is high levels of automation and standardization to reduce the number of humans involved. That said, this is a relatively small round of layoffs focused on a team not directly related to the product (recruiting), so it's not really proof that they are in trouble yet.
- bsbechtel 11y agoI agree that this may not be a sign of trouble, although this has been happening with more frequency across the on-demand sector. However, I would also consider most on-demand companies as service companies, not product companies, and in doing so, you consider the base labor cost the same as your cost of goods sold. So the 80% driver cut for Uber/Lyft would be the same as selling a retail product at a 20% markup (assuming sales taxes and payroll taxes/worker's comp are equal, which they aren't).