Sure, it offers an effective recommendation algorithm and a nice interface. But the music it streams can be accessed on other platforms - digital music kind of exists everywhere. Presented with a choice, most fans would rather listen to good music on a bad platform than lame music on a cool one.

Entrepreneurs, economists and VCs love the power of online platforms with network effects. This has arguably been The Big Idea of the past twenty years. But how much of a moat can any digital service have if the switching costs are minimal for producers and consumers alike? Fans of Taylor Swift can find her music on virtually every platform and the music coming through their headphones sounds identical. I'm sure there are Swifties who prefer Spotify over YouTube, but if she told them to switch, they would. Music platforms have a hard time differentiating themselves or exercising pricing power because their customers aren't really their customers.

This issue doesn't get enough attention, even though it applies to most digital platforms. Uber is so popular that its name has become a verb, but it still can't turn an operating profit. Rideshare is a commodity. I have multiple apps on my phone and take whichever is more economical. Many drivers do the same. It would be foolish not to, and there are third party apps that help both riders and drivers find the best option.

The stock market obviously disagrees with this assessment and acts like these companies will become reliably profitable any day now.

But Spotify was founded way back in 2006 and Uber is 14 years old. Any day now is getting a little long in the tooth.

Social media platforms have proven the exception to this phenomenon, but for a sinister reason: they've engineered their services to be highly addictive.