Granted, Ballmer made the mistake of putting Terry Myerson, who headed up the failed Windows Phone effort, in charge of Windows but that's another story.
You also need a plan. How would Windows Phone displace either Apple or Android?
If Windows phones would have had an emulated x86 mode, many people would have bought it instantly due to the momentum that now steadily decreases.
There can be solid business revenue if you are "just" #3 and the experience with development is very valuable. Although it is true that Microsoft and hardware has always been turbulent, with partners or without. Sometimes they simply created the best products in their class with a lot of margin, sometimes they basically sold scrap.
Even now, Microsoft has a great advantage over Google and Apple in getting meetings with the procurement people in those organisations.
Had they made it open source, it would have been a different story with Android and Windows Phone fighting to win the OEMs.
But that ship has sailed. Unless there's a paradigm shift in smartphones (doubtful), we're stuck with Android and iOS for the foreseeable future.
That would have necessitated open sourcing Windows
And then further the phone was a distraction for all of the other teams who were expected in someway to provide some software that would work on there as well as android and iPhone.
I agree that the phone would have been great … at some point. But in an MBA world, it was a liability
It's also a bit strange that the success of Windows was based on the ubiquity of clone PCs rather than single vendor, yet Microsoft instead tried to follow Apple here and let Google become the "clone PC" (Android phone) OS supplier.
I can't fault Balmer for at least trying to get a slice of the pie by belatedly putting out me-too products like Bing and Azure, but it's precisely because of Microsoft/Balmer missing the importance of the internet that it was put in position of being follower rather than leader.
Microsoft is really a bit like Intel in having totally dominated a product category, but then having missed on most of the major industry trends they might have been expected to lead on (for Microsoft, internet, mobile and AI; for Intel mobile, gaming and AI). They are lucky to have had a second chance with Nadella who seems much more in tune with industry trends, willing to rapidly pivot, and who seems to have made a masterful move with their OpenAI partnership in buying time to recover from an early lack of focus on AI/ML.
I thought it was a bad mistake to bow, kneel and surrender the smart-phone market space. Today, I am fully convinced it was a critical, life-threatening mistake as more folks move to the Apple ecosystem - buying both iPhones, Macbooks and Apple Watches because of a fully-integrated ecosystem. The funny thing it was Microsoft who popularized Continuity, but after they gave up due to lack of willpower, it was Apple who took over, executed better and won. Really frustrating to see the state of Windows OS and device market today.
I don't think they missed AI boat. Their culture would not have allowed them to create OpenAI, but they were fast to leverage their moat and push AI into their office and windows suites and azure. Hell, they are even trying to catch up with search using AI and are trying to push Azure for various AI startups and stuff.
MSFT rarely leads on anything - arguably even Windows is something they created being inspired by something else, while not going deep into hardware. Which what became the undoing of IBM. They are much better at being second. Azure - they were behind AWS, but not as late as Google.
I bet with Satya, even with mobile they would have grabbed Nokia much earlier and pushed Windows Mobile before Android took off.
AWS missed AI boat though.
I don't know the contractual basis of the relationship, but it seems this has to be pretty long term and strategic. A significant part of the competitive advantage of one AI vendor over another comes down to inference cost, which in turm comes down to customizing the model architecture for the hardware it is running on, which in this case either is, or will be, Amazon's home grown Graviton processors.
I don't know if Satya predicted it or not, but their push into open source and Github acquisition were very helpful for AI.
I'd be pretty intrigued but they're still struggling to nail the tablet market imo.
Google realized that if they want to stand a chance in catching up to the iPhone, they need to shove android in people's faces, and lure in devs.
Microsoft entered the game (WP8) when android already had a foothold, making it even harder. They started with a mostly empty app store, and while they were clever enough to make sure the most widely used apps would be available by effectively bribing those big companies to develop windows phone apps, they pretty much gave the middle finger to all the small indie devs. I remember when android 2 was around I just downloaded android studio and played around a bit, making a simple scrobbler app for my Samsung device. Sideloadong was king back then, but even up to this point I had to pay zero bucks and jump through no hoops to try this out. I don't remember what putting this on the Google play store would've cost me back then, but not much.
The windows phone experience was: sign up for a dev account to download visual studio with WP support. Start up VS, asked for your account again. I think in the beginning this was actually a paid account, probably because apple did it that way and again, you're Microsoft so act like you already own the place. But later in they reversed course here at least and you could log in with a free account.
So you start building a small test app and then you want to run it in the shipped emulator but surprise! Your laptop only shipped with windows 8 home which doesn't include virtualization features, so tough. So the only way to test the app was to push it to your phone, which was another overly complicated mess where your phone had to be in developer mode and you could only "sideload" one app at a time, iirc. The result was an app store with mostly tumbleweed. Whatever small utility or gimmick you wanted, when on android a search would give you dozens of results, on WP, there was maybe 4, and 3 of them almost unusable and abandoned.
I'm not blaming ballmer for having decided this specifically, but holy hell how did this pass any meetings with the higher-ups? You're uo against two tech giants who have a head-start of a few years, you try to get people to switch to your platform by being pricey, having no apps, and being hostile to smaller devs?
The same played out with all the phone makers, who had to pay license fees for WP when android was free to use. Guess which phones were cheaper in the end. And when Microsoft bought Nokia, Nokia had the unfair advantage of getting WP for free, making it even less attractive for others to compete in that sector.
And let's not get into the botched Nokia acquisition because I also don't think this can be blamed on ballmer that easily, or primarily.
This is something that Microsoft still struggles with. Some things have improved, but a lot of the dev experiences on the Microsoft side are still cumbersome and not aimed at small time devs.
My experience here is with browser extensions and publishing these for both old Edge (pre chromium) and the newer Edge. The entire publishing dashboard is/was overly complex and assumes you are either a single person or a big organization with (azure) AD set up. With Mozilla AMO you can just add individual developers to your extension by mail, and with Google it is as easy as setting up a group.
With browser extensions specifically (and Edge as well) you can also clearly see where it is still a dedicated motivated internal team setting things up and where things were handed over to more general teams and support was also outsourced to somewhere else.
Anyway, my main point is that even now, many years later, Microsoft still struggles in this area making me think this is more fundamental to the company culture and way of operating.