[JohnHaugeland]

I’ve been hearing people say this since this incorrect essay was written almost 20 years ago

Microsoft had 74% of the desktop then and has 72% today

Microsoft had 0% of hosting then and has the world’s second largest share, 23%, today

Microsoft had, in 2007, 18% presence share in console gaming. Today they have 65%

They were one of the four tech orgs present at the presidential inauguration

They have grown enormously in the time this essay claims they were dying

[hajile]

Xbox One sold 57.9M and Series family has sold 28.3M by mid 2024.

PS4 sold 117M and PS5 sold 61.7M by mid 2024.

Nintendo claims 150.9M Switch consoles sold as of the end of 2024.

That puts Microsoft at around 20.7%, Sony at 43%, and Nintendo at about 36.3%.

Even if you exclude Nintendo with a "no-true-gamer" fallacy, Microsoft still has just 32.5% of the market.

Windows was over 90% in 2007 and is just 72% today. If you include the super-important mobile market, Windows actual marketshare is something like 10-15% and even less if you include servers (where even most Azure servers run Linux).

[JohnHaugeland]

I said presence share, not market share

Three total customers exist worldwide. All three have red square, two have yellow square, and one has blue square

Blue square sold one in six items, so it has one sixth market share. Blue square is one in three households, so it has one third presence share

[pjmlp]

Yes, if we are counting only game consoles, now if we count game studios owned by each company, and they earnings across all platforms, it is a complete different picture.

Microsoft owns PC, and even Valve has forced to emulate Windows/DirectX to have any games on SteamDeck.

The amount of office worker typing Excel sheets on mobile phones is rather tiny.

[hajile]

Even if you look at just computer games, Windows has a smaller share than any time since the 1980s.

If you include the massive mobile gaming market, Windows gaming is an even tinier percentage of the overall market (maybe even <10%).

[pjmlp]

Have you ever played Windows games on the 1980's?

That would be a first.

What matters is where money is, and how much of those games trace back to Microsoft owned studios.

Good example with mobile games though, as it is a good example of Valve's failure to capitalise on 80% of mobile games being run with OpenGL/Vulkan, on a Linux like platform, and yet they have to translate Win32/DirectX, as means to get games on SteamDeck.

[JohnHaugeland]

The PC’s share of gaming, according to EA, by dollars spent, is 23%, the second largest slice behind iOS, larger than any of the consoles

Guesswork is rarely helpful

[PetitPrince]

> Microsoft had, in 2007, 18% presence share in console gaming. Today they have 65%

Source for this ? Also what do you mean by "presence" ? I had the impression that Sony was in a way better position than Microsoft, and they were both dwarfed by Nintendo by a substantial margin.

(not that this invalidate your overall point).

[pjmlp]

PlayStation 3 went down quite badly among game developers given how hard Cell was to program for, and users with the whole OtherOS fallout, this brought XBox 360 into the winning round.

Microsoft lost the plot afterwards with XBox ONE against the Playstation 4, and even with the Series S|X, they never recovered on the hardware side.

However this is kind of relative now, even if they don't public admit yet, the hardware is gone, they are going SEGA, and by being one of the largest game publishers, it hardly matters if the XBox console isn't going that well.

In one year they already recovered all the money lost in the ABK deal and litigations.

[jasode]

>I’ve been hearing people say this since this incorrect essay was written almost 20 years ago Microsoft had [...list of significant statistics...]

Whenever someone writes a provocative article about something being "dead", they are almost always talking about influence and mindshare -- rather than business statistics.

Yes, Microsoft is still a huge behemoth being a $3+ trillion cap company with a overwhelming marketshare of Windows & Office installations but the apex of their "industry influence" was the 1990s during the "Wintel" days before the internet came along. That 1980s/1990s was the time period when Bill Gates was CEO and "everybody was scared of Microsoft". Since, then they ... lost the browser wars (both old IE and new IE with Trident engine failed), lost the mobile shift (Windows phones failed), is a distant #2 in search engine market. Microsoft is somewhat back in the influence game with AI but that's because they partnered with OpenAI rather than build something internally. Arguably, it's Meta that gets more noise with LLAMA, and China's High-Flyer getting everybody's attention with DeepSeek-R1. That's the type of "alive vs dead" PG is writing about.

The "dead" being a writer's rhetorical flourish rather than a business status is the same when applied to "IBM is dead". In pure business metrics, IBM is still a giant company with $65 billion in revenue and $7 in profits. The airlines, major banks, and credit-card companies still run millions of transactions through IBM Z mainframes. Companies are still buying and upgrading expensive new Z mainframes. But the rhetorical "dead" means IBM's apex of influence was 1960s & 1970s. The later IBM trying to relevant with the newer tech like Watson and blockchain service doesn't matter to people.

Maybe writers should stop using "dead" as rhetorical technique because it just confuses readers. E.g. saying something like "DirecTV is dead" makes people scratch their head when they just watched a game on the satellite service last night. How would that be possible if it was truly dead?!?

[donny2018]

You are missing a lot of stuff Microsoft is doing. Azure, .NET, server tools, databases, VS Code, TypeScript, GitHub, (yes, OpenAI), gaming, XBox, desktop, business tools, Surface, Microsoft 365, Teams and lots more. I'd say much of the things they are doing is quite "fresh" and it's more relevant as it has ever been.

There is a reason it's market cap is bigger than Google's and Amazon's, and its downfall has been long overturned.

>with a overwhelming marketshare of Windows & Office installations

It's interesting that you mention it, as none of these are very important on their own to today's Microsoft if you check their latest quarterly reports.

[jasode]

>You are missing a lot of stuff Microsoft is doing. Azure, .NET, server tools, databases, VS Code, TypeScript, GitHub, [...] and it's more relevant as it has ever been.

I didn't list them because they're not "relevant" (scare quotes) to PG's rhetorical angle of "dead". Yes, of course those Microsoft components are still relevant and still being updated and modernized. That said, even though I personally use VSCode, Visual Studio, Github every day, and have upgraded too many MS SQL Server databases... my point is those examples of Microsoft's current usage is not what PG is talking about. I'm not saying readers have to agree with PG. They just have to understand that he's using "dead" as a provocative shorthand about "influence" rather than business stats.

Same confusion as IBM coming out with new Z mainframe models in 2025 and IBM Red Hat just released a new RHEL 9.5 a few months ago and yet people will say "IBM is dead". How can IBM be dead if Red Hat Linux is still relevant?!? That's the problem with different readers' interpretation of the word "dead".

EDIT reply to: >Then what is he talking about when he says "dead"? [...] I mean for vast majority GitHub is a synonym for Git and VSCode is nearly a de-facto IDE for frontend development,

Github (2008 acquired by MS in 2018) and VSCode (2015) didn't exist in 2007 when PG wrote his "Microsoft is Dead" essay. It's possible those are "influential" enough to change his opinion. Maybe not. The examples of millions of people using MS Excel and Word every day back in 2007 with no meaningful competition from Google Docs or LibreOffice didn't stop him form writing "Microsoft is Dead". Therefore, we must conclude he's using "dead" in a very particular way.

[vishnugupta]

> VSCode, Visual Studio, Github every day, and have upgraded too many MS SQL Server databases... my point is those examples of Microsoft's current usage is not what PG is talking about....

Then what is he talking about when he says "dead"?

Also comparing those MS softwares with Z-mainframe & RHEL feels a bit off. If you take a 90th percentile of s/w developer starting career today they are more likely to have heard or used those MS tools than IBM's. I mean for vast majority GitHub is a synonym for Git and VSCode is nearly a de-facto IDE for frontend development, TypeScript I don't need to say much.

[JohnHaugeland]

I understood the writer's flourish.

If you want to understand my reaction, take two steps:

1. Note which companies do in fact have mindshare

2. Check which of those are owned by Microsoft

[rcxdude]

It is a pretty unhelpful way to say "Is not longer as incredibly dominant as it was". It's a stretch to apply it to "Is declining into irrelevence", even.

[nunez]

Watson has been around a long, long time, and they are still leaders in quantum computing IIRC.

[dijit]

> Microsoft had 74% of the desktop then and has 72% today

Microsoft had 90%+ of desktop penetration in 2008; in fact, it made news that it had slipped to below that at the end of 2008.[0][1]

Now it's around 70%, but seems to be improving?[2]

[0]: https://www.osnews.com/story/20605/windows-market-share-slip...

[1]: https://www.computerworld.com/article/1367310/windows-market...

[2]: https://www.statista.com/statistics/218089/global-market-sha...

[Clubber]

>Microsoft had 74% of the desktop then and has 72% today

That's only counting the latest version, if you include all Windows versions, it was in the high 90%. Today it's in the mid 90%.

https://gs.statcounter.com/windows-version-market-share/desk...