> Apple publicly stated they expected to sell 3 million of the things in the first year.

If they did, I'm sure that OP or you can find a source for this statement. A rumor published by Bloomberg (or, in a more charitable interpretation, an estimate given on background) is not "publicly stated".

> As the article states, Apple publicly stated they expected to sell 3 million of the things in the first year.

This is false. The link goes to a Mark Gurman article in Bloomberg that claims Apple at one point had an internal prediction of 3 million, and that it was revised down before it went on sale. That may or may not be accurate reporting, but it’s not close to claiming they publicly predicted 3M sales.

When the iPhone came out, it was perfectly positioned. Smartphones weren’t new of course, but they were clunky devices with tons of compromise. The iPhone was designed to fix a lot of the problems around existing smartphones by having great built in apps, a great browser, and great usability thanks to many novel and intuitive forms of interaction. When it came out it was a must-have device and it solved so many problems for so many people.

For everything I’ve seen of the Vision Pro it has represented a solution in search of a problem.

Also worth noting they sold about 270,000 of them during the first week of sales, which is about $1 billion of revenue in just the first week. I think if they didn't over-manufacture them against unrealistic expectations that they may have gotten a reasonable profit, or broke even on paper (at least, ignoring their IRR baseline). $1 billion is enough to pay 500 engineers $300,000/year for 6 years.

But I'm sure the cost of components ate a ton of that, and I'm not sure how much of that would be fixed vs. marginal costs. For something like the custom silicon, I could imagine fixed costs were substantial compared to marginal costs. I'm also not sure how much of this can be amortized across future versions of the Vision Pro, or if a huge amount of the fixed costs will need to be repeated.

Maybe for the amount of silicon fabrication tooling, and the complexity of manufacturing tooling for something so unique, the fixed costs required the sale of 2-3 million units to break even. If so, I could see various parts of the organization feeling pressured to deliver analysis which confirmed the necessary level of demand, especially if executives expressed bullishness on the project.

When does apple ever publicly predict sales?