The article claims, if I understood it correctly, that GCP will not be profitable unless it reaches the scale of being at least #2. The implication is that economies of scale are necessary for profitability.

For example, in 2014, James Hamilton spoke about how every day that year, AWS was adding enough new server capacity to support all of Amazon's global infrastructure when it was a $7B annual revenue enterprise (2004). That scale of build-out, server purchasing, etc. presumably justifies and enables significant cost reductions, such as buying equipment from vendors in bulk with steep discounts, or buying or building large facilities at once, etc. AWS is custom designing many parts of its stack including networking chips and software, ARM-based CPUs, and more; that kind of investment is only worthwhile at a certain large scale. This scale presumably motivated AWS's purchase of chipmaker Annapurna Labs in 2015. Some operator of a few colocation facilities, by comparison, could not justify designing custom silicon since their scale would not allow them to recoup the R&D costs. Before AWS reached whatever scale made these kinds of investments worthwhile, they were purchasing commodity hardware.

To pick another example, a company like SpaceX can only be profitable if it launches a certain minimum number of rockets per year. Without enough rocket launches, and demand from customers, you can't recoup the R&D budget needed to design them.

Google has a head start in cloud because their company's existing facilities are already huge. However, public infrastructure has different demands than internal infrastructure, and it sounds like the necessary investments in public have been costly.

But there's a subtle difference between being big enough to make profits and ranking. If you need X amount of business to make profits, why does it matter whether one or two guys make X+ ?

If you take your SpaceX example, they need to launch some number of rockets. So what if someone else is launching more?

Why isn't the minimum expressed as a % of the market for example?

No.

No CTO who valued their career would go with GCP. “No one ever got fired for choosing AWS/Microsoft”. Even if Azure isn’t held in as high esteem as AWS, businesses trust Microsoft.

it did! an entire generation of MBAs learned from jack welch, who popularized be #1 in anything you compete in (and if not, exit).

for those interested in management fads, https://www.investors.com/news/management/leaders-and-succes...

We also have Jack Welch to thank for stack ranking (ranking the bottom X% of employees in any given team, no matter how elite they are, as "underperforming").

The self-inflicted damage management fads (and Welch's Ilk in particular) have caused to our economy is incalculable.

Wasn't he also the person responsible for "building up" GE Capital, i.e. the division that almost took down GE in late October - early November 2008?

They have to fight even for the #3 spot. They're currently losing to Alibaba!