3 ms·
There's a talk a couple years old now by an Uber engineer called something like "what I wish I knew before scaling to 10k microservices", I haven't watched it i
by throw__away7391 4y ago
There's a talk a couple years old now by an Uber engineer called something like "what I wish I knew before scaling to 10k microservices", I haven't watched it in a while but when I saw it I remember thinking "these people are absolutely insane". I've since heard some crazy stories about things like new services being built because people would rather build a new service than talk to the devs that owned the existing one. I don't know if this is true or not or how widespread this was, but the impression I get is that Uber has massively over-built in an effort to look more like a "tech" company rather than a "taxi" company to investors.
We have a couple of local companies in NY with their own ride-sharing apps. They aren't as polished as Uber, but they do work and the companies that built them have about 1% the staff of Uber.
- sokoloff 4y agoIf Uber is covering 200x the market globally as a local NY company that has 1% the staff, it seems like Uber is more efficient on a sales/employee basis. I doubt NY is 1% of the phone-hailed taxi market globally.
- throw__away7391 4y agoWell sure...but the thing about software that's different from say taxis or commercial real estate is that once you build a product the marginal cost to scale it up is close to zero. This is what drives the valuation multipliers of growing tech companies. If your software requires 10x the engineers to support 10x the usage I'd say such valuations are not justified.
- monocasa 4y agoParticularly something like Uber where there's a natural sharding that can happen at about the scale a regular taxi business would service anyway. There's some overhead in when payment methods start changing, and complying with local laws in a new metro, but it's pretty damn close pure geo sharding.
- sbierwagen 4y agoThe point of the tech industry's relatively high valuation is that headcount isn't supposed to linearly scale with sales! If you need to add one employee for every 100 customers, then you're not Microsoft, you're McDonalds. If Uber has to staff up a hundred person office in every country they operate in, this is bad, and guarantees they won't ever be profitable!
- deepsun 4y ago"About 24 percent of Uber’s bookings—all the money that customers pay through the app and in cash, including driver earnings—occur in just five cities: New York, Los Angeles, San Francisco, London, and São Paulo." 2019