I was picking up food the other night on Bloor Street (via Uber Eats) and the lineup of delivery drivers outside of the restaurant was at least ten people deep when we arrived. While we were waiting, another handful of drivers pulled over to quickly pickup their deliveries. This is what is happening in our cities right now, especially here in Toronto while we live through another stay-at-home order. And the numbers certainly reflect it.

Last month in March, Uber's delivery business (which is separate from the company's mobility business) recorded a 150% year-over-year increase in annualized gross bookings. The company's run-rate as of March is now $52 billion. To put this number into perspective, the company's mobility business also had its best ever month in March with an annualized gross bookings run-rate of $30 billion.

Delivery > mobility right now. Makes sense.

To further put this into perspective, total restaurant spending across the entirety of the United States was $670 billion in 2019 (figure from Benedict Evans). So Uber Eats has quickly become a meaningful part of how we eat. I obviously believe that people are dying to get out and eat at restaurants again, but these figures are still interesting nonetheless.

It's also interesting to think about the above trendline from a broader logistics perspective. Alongside the rise in Uber Eats, we are seeing a wave of capital move toward "rapid delivery apps." These are platforms that allow meals, groceries, and other stuff to be delivered, in some cases, almost right away, which aligns with where I think consumers are moving. Rather than making lists and doing weekly shops, it's now about just-in-time delivery.

It's arguably a lazier way of going about things, but water will always find the path of least resistance.

Many, or perhaps most, of these platforms have adopted an asset light approach. Instacart, which partners with existing grocers, would fall into this category. Their model revolves around gig workers going into existing stores, picking orders directly from the shelves, and then delivering those orders. And it is what Blair Welch was getting at in his recent RENX interview when he reasoned that grocery shopping is still being done, almost exclusively, at local stores.

This approach is enough for Instacart to be valued at nearly $40 billion, according to the Financial Times. So something seems to be working.

Amazon is sometimes criticized for its private labels. The way this generally works is that Amazon uses the data that it collects from its platform to see what customers are buying. It then goes out and makes its own version of these products and sells them in competition with the other products in its marketplace. The reason why Amazon (and others) do this is because the margins are generally better on private labels, even though they are often positioned to the end customer as being a value-oriented alternative. That is, they're cheaper.

Some people think that Amazon shouldn't be doing this, particularly as its third party marketplace continues to grow. This side of its marketplace deals with inventory that Amazon doesn't own. It is the stuff of third party sellers who come to the platform to access Amazon's customer base and reach, and to possibly use its fulfillment services. This marketplace now makes up about 60% of Amazon's sales volume and so it has become a dominant part of its business. It's a way to grow without having to spend money on additional inventory.

Is it, then, acceptable for Amazon to mine this data, replicate products, and compete with its own customers? The truth is that this isn't all that new. As Benedict Evans points out in this recent post, retailers have been doing this for more than a century. The above table taken from a 1932 report on "chain store private brands" shows that about 80% of stores in the US at this time were selling private label brands. Furthermore, it represented about a quarter of their overall sales. Is this time any different?

Table via Benedict Evans

Every year, Benedict Evans publishes a presentation about the "big macro tech trends" impacting the global economy. They are always excellent and I usually share them here on the blog. It's also becoming harder and harder to differentiate tech trends from the rest of the economy, and so in many ways this is just a presentation about important macro trends.

In this year's presentation, he focuses on the "unbundling" of retail, ecommerce, advertising and TV; China and the end of the American internet; and a few other timely topics. To view the presentation, click here. Benedict also delivered this same presentation at a recent event by Protocol and Nasdaq (video link) in case you'd prefer to consume the content that way.