How the Mighty Have Fallen

According to Paul Graham of Y Combinator, Microsoft is Dead:

A few days ago I suddenly realized Microsoft was dead. I was talking to a young startup founder about how Google was different from Yahoo. I said that Yahoo had been warped from the start by their fear of Microsoft. That was why they'd positioned themselves as a "media company" instead of a technology company. Then I looked at his face and realized he didn't understand. It was as if I'd told him how much girls liked Barry Manilow in the mid 80s. Barry who?

Graham was quick to clarify in a follow-up message that by "dead" he didn't mean dead in the sense that the company is about to collapse, but in the sense that Microsoft has lost its industry influence - it has ceased to matter as a factor for other software companies when deciding on strategy.

Ceasing to matter doesn't mean a company is going to go out of business next year, any more than it means a pop star will suddenly become poor. But it probably means there is trouble ahead. Actors and musicians occasionally make comebacks, but technology companies almost never do. Technology companies are projectiles. And because of that you can call them dead long before any problems show up on the balance sheet. Relevance may lead revenues by five or even ten years.

I think he's absolutely right.

Microsoft is the "village elder" of software companies. Many old, grizzled players in the market think that Microsoft is still a key factor in decision making, and continue to believe (and repeat) the message that they are the number one choice for business software. Meanwhile, the next generation of technology businesses have sprung up and completely ignored Microsoft technology, creating their own software solutions or adopting innovative new applications from other new companies.

Microsoft still seems to think it can afford to run a business model of waiting to see what other products are successful and then either buying or copying them late into the game, instead of innovating - the so-called "Embrace and Extend" strategy.

They have of course always worked this way. DOS was bought by Gates at the start of his entrepreneurial career, and then later Windows was copied from Apple's Mac OS. When the Web took off, Microsoft beat Netscape to a bloody pulp by matching it feature for feature whilst intentionally introducing incompatibilities in the way it used JavaScript so that developers would be forced to pick a side. The final nail in the coffin was when they bundled IE with Windows so users wouldn't realise they had a choice.

But more recently they have stumbled. Despite their vast resources, Microsoft is now so far removed from the front lines that by the time an idea worth stealing has reached their attention it has already dominated the market. This is not a sensible time to jump in, especially if you have no idea what makes a good product and resort to trying to sell a poor knock-off at the same price point with nothing positive to differentiate it.

The iPod was around for five years before it caught Microsoft's attention. Over that time it has been refined over several generations into a line of several well-designed products. The Zune by comparison is an uglier rip-off of the original 'pod with a smattering of mostly useless new features, and some poorly-conceived, incompatible software to drive it.

And what about Google? Google came along and stole the search engine market before Microsoft realised there was such a thing (they presumably thought everyone was still using AOL keywords or something). Five years later we get the new improved MSN search - now almost as good as Google, and it's from Microsoft! Yeah, that'll make it a hit.

And of course that's the key symptom of their demise - "made by Microsoft" is no longer a selling point. If anything it's a detractor. Whereas once it might have meant "industry standard" and "compatible with everything" now it means the opposite. Now that we have reliable, free, open-source products based on actual written standards, why would we want to rely on half-baked, proprietary, single-vendor garbage that will probably be dropped or changed beyond recognition in the near future in accordance with the whims of Microsoft's corporate strategy.

But while Graham has in many ways hit the nail on the head with this article, as I read his reasons for Microsoft's decline, I realise that I must disagree quite strongly with some of the conclusions he has reached:

He cites Google as Microsoft's biggest competition now, and I think there's some truth to that, but I think his reasoning about why has been pulled from thin air:

When did Google take the lead? There will be a tendency to push it back to their IPO in August 2004, but they weren't setting the terms of the debate then. I'd say they took the lead in 2005. Gmail was one of the things that put them over the edge. Gmail showed they could do more than search.

Gmail? Seriously? I wonder what proportion of Google's total user base has a Gmail account? One percent? Probably less. I would guess also that Hotmail is much more heavily subscribed. And what about instant messaging? Where is Google Messenger?

No, on the contrary, e-mail and instant messaging is one of the few Internet arenas in which Microsoft still dominates, and rightly so since Outlook and (the Windows version of) MSN Messenger are some of their finest consumer products (though I daresay that open-source offerings like Thunderbird and Adium have got them on the run). Even humble Hotmail has existed as the premier web e-mail service since long before Google appeared on the scene. All Gmail did was give it a lick of paint and some fancy special effects (which, in the process made it incompatible with all but the top two or three browsers).

Gmail is neither the most popular (Hotmail) nor the most technically sophisticated (Yahoo) webmail offering on the market, and far from proving that Google can compete with desktop apps like Outlook, it leaves me wondering "is that the best they can do?". It also opened the lid on Google's security policy. Do you really want a giant corporation knowing everything you write, everything you like to search for, and everyone you know?

The thing that proved that Google could beat Microsoft was advertising. Google's idea of using search data for targeted advertising was brilliant, and providing their ad services in the form of easy-to-use, free development tools for bloggers was an absolute master stroke. All those years that Google was growing as a search provider Microsoft was no doubt laughing at them an wondering how they ever planned to make money. Well now they know, and if rumours are true they are scrabbling as we speak to purchase Google's closest advertising competitor, the (far inferior) Doubleclick (*) so that they might have a hope of competing.

What is wrong with Microsoft is that they don't know what they are good at. They seem to be labouring under the impression that they are good at everything. In fact they are only good at a few things, and they've now spread themselves so thin that they have let their core businesses begin to stumble. Microsoft Word 1.0 was good. But every subsequent version has added more bloat and useless features so that users are now crying out for a superior alternative. It's pretty good going when your product is so bad that users are actually looking to buy a competitor's product even when none currently exists (Open Office notwithstanding).

Microsoft can't do chic, so there's no point going after iPod and iTunes. They can't do web applications so there's no point going after Google or Apache. They've let their browser languish so long that I believe they would be better off following Apple's example and switching to an open-source core. Trying to fight against these realities will just cause them to burn more of their monetary and creative resources, and attract mockery at the poor quality of their offerings. I don't think Graham quite gets this - this in particular made me laugh:

Gmail also showed how much you could do with web-based software, if you took advantage of what later came to be called "Ajax." And that was the second cause of Microsoft's death: everyone can see the desktop is over. It now seems inevitable that applications will live on the web - not just e-mail, but everything, right up to Photoshop. Even Microsoft sees that now.

The reality is that Bill gates "saw" this concept before anyone else. Back in 2000, Microsoft was arguing that the future of software was going to follow the rental rather than ownership model. The idea was that rather than buying and owning software, we would all rent our applications (from Microsoft) via the Internet, paying by hour for services. Documents too would be stored safely in giant Microsoft server farms where they could be backed up (and no doubt automatically spidered for child porn or other illegal material). For Microsoft, this dream of returning to a world of supercomputers and dumb terminals would be ideal. Since they aren't a hardware vendor, Microsoft would only stand to gain, while competitors like Apple and IBM who rely on hardware installation profits would be quietly destroyed - an added bonus.

But Microsoft discovered that it didn't actually have the clout to effect this kind of industry change. They tried a few tentative steps towards this, first by merging IE and Windows so that their desktop OS effectively became a web browser, and then by sticking ".net" on the name of all their products, and re-jigging their software architecture so it would be better suited for client-server delivery and cross-network inter-application communication. But in the end nobody bought into it. The kind of web-based micropayments needed to fund this model have yet to appear, and more and more software has become available for free, making the idea that people would pay hourly rates to rent access to applications seem ridiculous.

In the end I think this idea has failed partly due to Microsoft's technical incompetence, but primarily because as a concept it really sucks! People want more privacy, not less. And most people prefer to own something rather than rent it. Compare the success of iTunes versus Napster-like "MP3 rental" systems. People will often actually pay more overall if it means that they can say that something is theirs.

The Web is not a great platform for all application types. For networked, or shared data it makes sense. People who want to share or store their photos have flocked to sites such as Flickr, and social networking - communication between friends-of-friends - has really taken off. But tasks such as media playing or single-player gaming are performance-limited. And why would you want to download all the resources for a game every time you play it, instead of just downloading/installing it once and playing it locally? Caching helps, but then that would have security implications for "pay per view" software. It all just seems like a solution looking for a problem.

Adding the latency of network transfer of all personal documents before they can be played or edited doesn't make sense, and if you don't transfer the documents then you have to do the bulk of the processing at the server end, which means replacing our existing distributed processing network of desktop PCs with giant monolithic supercomputers that have to service the needs of hundreds of users. Who is going to pay for those? Or pay to rent their services when it is easier and cheaper to put the same power in a box on your desk? Bandwidth and processing power are probably going to be the key resources of the future, and using powerful desktop PCs as dumb terminals seems like a particularly foolish way to squander them.

So Graham's claim that Microsoft was killed by Gmail and broadband Internet is nonsense. If anything these should have given substance to Microsoft's vision of a client-server model for business software. But the fact that they haven't tends to confirm my hypothesis that the idea was ill-conceived, and that Microsoft was already floundering long before these products came along (it just hadn't realised it yet). It's sad that the one time it had an original idea, other companies came along and made it reality (albeit an impractical, toy reality) long before Microsoft could shift its momentous weight far enough to start bringing it to market.

Graham thinks that Microsoft's problem is inability to code, rather than lack of focus. And his proposed solution is for them to simply buy in more talent:

The surprising fact is, brilliant hackers - dangerously brilliant hackers - can be had very cheaply, by the standards of a company as rich as Microsoft. They can't hire smart people anymore, but they could buy as many as they wanted for only an order of magnitude more.

But Microsoft has very smart people already, the problem is that it is wasting them on derivative crap like the Zune or MSN search instead of giving them something useful to do. The only worthwhile part of Microsoft right now is the team that built the XBox 360. Those guys really know what they are doing. They've created a great product that consumers want, they've made it cool, they've marketed it well, they've delivered it on time and they've actually had some original ideas (such as XBox live).

I picture the Microsoft XBox team as being very reminiscent of the group that originally built Microsoft Word, or the first incarnations of Internet Explorer. They are working on something fun and exciting, with just enough industry competition to keep them on their toes, but with goals that are achievable, and a very real chance to do some things that have never been done before.

Just imagine by comparison how soul destroying it must be for any talented developer to work on the next version of Windows or Internet Explorer. Your brief is to modify part of a trillion-line code base that was originally written using practices that are now fifteen years out of date. You can't fix certain bugs because you have to retain compatibility with earlier workarounds. The project is too large for any one individual to envision the whole, and the aim of the project is not to develop any new features, but simply to copy the top ten ideas from your nearest competitor, who already released their product several years ago and is now publicly mocking you for every day that your project is delayed.

Vista was a ten-year project that resulted in an OS that almost (but not quite) matched the feature set of Mac OS X, released six years beforehand. Even if you believe Gates's ludicrous claims about who copied who, that still means that Apple can take MS's ideas and implement them years before Microsoft can get its own product to market.

I don't believe Microsoft has been killed because Google has got too big. On the contrary if Google getting big is a danger to anyone, it's Google itself. Microsoft has been killed by a thousand smaller, more agile companies each doing a bit of what Microsoft does, only better. Microsoft meanwhile are so bloated that they are unable to even maintain their flagship products at a level with fresh-faced competitors, let alone make innovations or set new standards. And they are so hated by their own customers that even the ones who continue to use their products laugh when they fail.

Buying up talent won't help Microsoft to recover. In particular I suspect that buying dozens of "cool" startup companies like Google and Yahoo have been doing is a sign that they are destined for the same fate as MS in the long run. An idea is worthless unless you can capitalise on it, and the modern rate of innovation is so high that the only way you can do that is if you think of it yourself. By the time you have noticed that someone's had a good idea, it's already too late to make any money out of it. Buying up trendy companies smacks of desperation.

*UPDATE: Google beat them to the punch it seems. Yep, Microsoft is toast.