If you look back at the creation of the PC in the 1980s, they were just desktop machines that sat in place. They were supposed to be portable but, for the most part, weren’t. Yes, we had PCs from gaming companies like Atari and Commadore, but other than some of us using them for schoolwork, they weren’t taken seriously. IBM PCs and their variants were heavy. The Panasonic I had weighed at a whopping 35 pounds. It came with a printer, and it cost me over $3K then, which would have been $8,514 today if you take inflation into account.
Eventually we got laptop computers, and they were pretty much made by the same people. You had desktop PCs and laptop PCs and sthen came the first smartphone from IBM called Simon (don’t worry; few bought it). After that, we had PDAs that came from a whole different class of company. Palm, Research In Motion (BlackBerry), HP and even Sony got in this market along with Microsoft. Apple, which started the PC revolution with the Apple II, had Newton. On paper it was more capable than the rest, but Jobs killed that and later brought out the iPod. It was just for listening to music, but it gained functions until it morphed into a phone. Over the next few years, it forced even massively powerful cell phone vendors like Nokia, who bet on Microsoft, to their knees.
So why wasn’t the Pocket PC the foundation for the smartphone, which is basically a pocket PC that can make calls? It had nothing to do with telephony and everything to do with execution. And I’m seeing similar execution from Microsoft on AI, suggesting it may be another Zune-like event.
Microsoft’s and Intel’s Recurring Problem
Microsoft doesn’t learn from its successes; it isn’t even clear it learned from its failures. I think that’s because Microsoft turns over key people, and those left in charge just don’t want to consider that their predecessors actually knew what they were doing.
When you look at the greatest product successes Microsoft ever had, (besides Office, which started as a productivity product for Apple), Microsoft has had two massive user successes: Windows 95 and Xbox (Azure is more of an enterprise play). Both used basically the same play book but were run by different people.
The formula for success was to put real marketing people in charge, not ex-military people or some reassigned engineer, but people that understood how to market and then give them the budget they need to be successful. That set up resulted in successes like Windows 95 and Xbox.
What was particularly funny about Windows 95 was that marketing did its job well, but operations did not, but Microsoft responded as if the opposite was true. Apparently, operations was better connected because when Windows 95 blew up largely due to poor code choices and horrid customer support, the conclusion was that demand was a bad thing, not poor execution. To this day it is hard for me to understand how they reached that conclusion.
With Xbox, Microsoft again put people who knew how to build demand in charge and took on Sony, a powerfully entrenched consumer electronics company. Xbox would have eventually taken out the PlayStation, but then Microsoft kicked out its lead people and cut the marketing budget. While Xbox is still a powerful platform to this day, it could have been far more. It wasn’t competition that killed it. It was Microsoft’s leadership.
Intel has gone down a similar path, but at least it hasn’t repeated the same mistake over and over again. Intel’s success bloomed under its best CMO, Dennis Carter, who turned an ingredient brand, Intel Inside, into a badge of quality. But he didn’t train his successors, so the amazing brand he built with Andy Grove’s massive support was weakened over time to a point where Pat Gelsinger, who had learned under Grove, also didn’t get the power of marketing and under supported it. I believe that led to his early retirement (termination).
So, had Microsoft executed the Pocket PC and the Windows Phone to the level they’d executed Windows 95 and Xbox they would have been a success, but they didn’t, and the execution around AI is far more similar to their failures rather than their successes.
Wrapping Up: Why Microsoft Is Likely to Fail with AI
AI is kind of like Internet Explorer in that Microsoft was late to the game but licensed a technology to close the gap (the browser was Spyglass; with AI it is ChatGPT) and brought it to market. But execution has again been lackluster. Microsoft had to pull back the strongest feature, Recall, because marketing didn’t position it properly. Apple brought out an identical feature, presented it better, and didn’t have to pull it back.
Both Microsoft and Intel need to up the marketing game significantly if they want to be successful going forward. Microsoft needs to pull from the learnings of Brad Silverberg and Rick Thompson, while Intel needs to pull from Dennis Carter to change its trajectory. Otherwise, both firms are likely to miss the AI wave rather than lead it. These skills are in the companies’ histories, they just need to re-learn them. And they need to learn that marketing isn’t optional. It is a requirement for success.
One final reminder. While I worked there, IBM nearly crashed and took massive brand damage, and IBM’s board fired John Akers, who was one of the most qualified CEOs on paper but who was done in by his staff. IBM didn’t hire an engineer or even pull its own people. Instead, it hired Louis Gerstner. Gerstner knew nothing about tech but was a brilliant marketer and he saved the company by first rebuilding its image and brand.
That’s the power of marketing, and too many tech companies forgot this after the 90s. It is also why Qualcomm’s AI PC effort didn’t reach its potential. The ironic thing is that AI is potentially able to recreate the skills of people who are no longer in the workforce, suggesting these companies could build the talent they need if they can’t hire it. Microsoft’s AI effort reminds me a lot of Zune, a lot of inside belief in success, but not enough outside execution, and I just don’t think it will end well.
