I'm not so sure. Apple hires the best and their evaluations of new candidates are rigorous, so I doubt they have idiots running things. We all know there's a lot of the worlds best talent at Apple. But they struggle a lot with the web services they've tried to create. Mobile me, iCloud, game centre, ping, Maps; these have all fallen short of Apples promises. It's a repeating pattern that to me indicates a culture issue inside Apple.

If it was a culture issue it would extend to the successful product teams as well. Those failures point to the people in charge of those groups, not the Apple culture.

Unless by Apple culture we mean giving people lots of responsibility and autonomy.. then, yes. But that's also the key to their successes.

Is it possible that a culture which is effective at producing "good" outcomes in one domain is ineffective at producing "good" outcomes in another? Should managers use the same techniques to manage steelworkers as they do software engineers?

Edit: steel works -> steelworkers

That's a good question, but in this case they're both software engineer domains. Apple does run some pretty successful online services, like the iTunes and app stores. If your theory were true they wouldn't be able to do that. Clearly they do have more problems with services than software/hardware. But that could be a function of the people at the top, not the management process, which but most accounts is minimal.

I'd argue that there are a few aspects of providing products and services for the web that are unique and not relevant to offline software development.

For example, scalability. Most of Apple's problems in this domain have been associated with scaling. iTMS and the App Store both scaled very well over time, but you have to consider the size of their audiences at launch. iTMS launched to a relatively small market of Mac and iPod users. The App Store launched to a relatively small market of iPhone users. Now that both the OS X and iOS user bases are significantly larger, Apple has struggled immensely to launch web services.

When developing for the web, your job is not done until you've released your product into the wild and figured out where the bottlenecks and weak points are with a massive user base. Apple's culture of secrecy doesn't facilitate this part of the process. Until that changes, they will be taking a huge gamble every time the release a new web service or improve an existing one. And, as with iClouds Core Data woes, developers who depend on these services are beginning to lose patience with Apple. This should be a major concern but it appears, from the outside looking in, that it isn't.

From what I know from a few engineers that were directly involved in the MobileMe launch, yes the failure was in the engineering management that did not want to listen to the engineers/project managers/qa folks, but instead wanted to please SJ by showing nice Ajax UI.

This is why the VP of MobileMe was let go soon after: he came from Microsoft/Starbucks, and did surround himself with various ex-microsoft folks.

If you know folks at Apple that were in that group at the time, you will find out that the failure did happen at the VP/Director level that didn't want to listen to the project managers / engineering leads / QA leads.

If the engineering management sucked, then that was also Steve's fault. You're not the iron fist of Apple and get all the credit without taking any of the blame.

Indeed, and that is why that management was replaced.