Why Google need not fear Microsoft

Some people are afraid that if Microsoft integrates search functions into Windows, it'll kill off Google - but the competition should be welcomed

As the world awaits the European Commission's antitrust ruling, certain pundits see the makings of a future investigation as Microsoft moves to compete more forcefully in Internet search, a domain where Google currently dominates.

The fear, in short, is that Microsoft will integrate aspects of search functionality into Windows in such a way that it will foreclose opportunities to Google. No company stands a chance when faced with such integration, leading to the eventual demise of Google.

To put it simply, I don't agree. Google, as a company which accounts for 80 percent of all Internet searches and a billion dollars in estimated yearly revenue, is a company deserving of competition, and perfectly capable of standing on its own.

First, recognise that a contest with Google would be one between two dominant companies, only one of which has extensive experience in search technology. Second, Microsoft's track record shows that the company competes more on technical merit than distribution. Last, Microsoft should do more to defuse antitrust regulators' fears by making a strong and public commitment to openness in the manner in which search functionality will be integrated into their products.

Microsoft, the critical competitor
How many companies would go up against the dominant Sony game system, PlayStation 2? Sony is the gorilla in the game console space, and combined with the installed base of PlayStation games (all of which are compatible with PlayStation 2), Sony's position seems particularly entrenched. Even so, Microsoft spent billions of dollars entering and competing in that market, posing the most credible threat to PlayStation's dominance.

The same applies to other markets Microsoft entered, many of which are dominated by a single player. Microsoft competes with Palm in handheld operating systems, Symbian (Nokia moved to take controlling interest of Symbian on Monday) in operating systems for "smartphones", Oracle in databases, and America Online in Internet access. Though Corel's WordPerfect these days is considered an also-ran, WordPerfect was the dominant word processing product. When Microsoft came out with Internet Explorer 1.0, Netscape owned a whopping 90-95 percent share of the market for Web browsers.

In short, Microsoft is one of the few companies to pose competition in markets dominated by one company. Few bother to challenge Google in Internet search technology, in spite of last year's estimated billion-dollar revenue haul. Microsoft is "bothering", serving the same important roles its serves in other markets, which is sole competitor to a dominant player.

To Microsoft's critics: An honest accounting
The subtext to the claim that integration is the way Microsoft has beaten past competitors is that Microsoft only succeeds by leveraging its dominance of desktop operating systems. Microsoft has the ability to integrate its products into Windows, giving those products theoretical access to 90-95 percent of desktop consumers.

Though perhaps true in a limited sense over the long term, it applies little within the time frames at which computing evolves. The integration advantage theory presumes that everyone upgrades to the latest version of Windows shortly after Microsoft releases it. That is far from the truth if usage figures for Windows XP are any indication. Only 42 percent of Windows installations currently run Windows XP, an operating system that has been available for over two years. This means that features integrated into XP hardly gain a distribution that is far above what is possible for other companies.

Sun has managed to get the Java runtime onto HP, Dell, and Gateway computers, who in total account for a 55 percent share of the US market (although only 33 of the worldwide market, but I haven't looked into whether Sun has made agreements with non-American PC manufacturers). That's higher, in other words, than the share Windows XP has of the total Windows marketplace. Furthermore, such an avenue of distribution (paying OEMs to include your products) is unique to Windows, as Microsoft doesn't make the hardware upon which its software runs, while Apple and Sun do (in other words, Microsoft won't be getting .Net pre-installed on Solaris anytime soon).

This means that almost half of those who have the latest version of Internet Explorer (which is IE 6.0, and has a 70 percent share of the Web browser market) acquired it by downloading from the Web, since IE 6.0 was only an "integrated" product on Windows XP. This is little different than the situation faced by many of Microsoft's competitors.

To put it simply, integrating products into Windows doesn't give quite the boost to Microsoft's products that its critics would have us believe.

Besides, you'd have to be pretty lazy not to download software these days. You don't even have to go to the STORE to get software, as was common when Windows 95 burst onto the scene. You can sit naked at your desk and buy a product by, effectively, pushing buttons. Does Microsoft really have such a distribution advantage in such a globally networked environment?

Likewise, the free distribution thesis glosses over the very real technical advantages Microsoft has imbued in products it introduces to new markets. As I've noted numerous times in the past, Internet Explorer was the FIRST browser to offer itself as a set of reusable and extensible parsing components. AOL continued to use IE even after they had purchased Netscape. They did this because Netscape COULDN'T offer the level of branding and customisability offered in the then-current version of Internet Explorer.

In fact, the Mozilla open-source development group implicitly admitted this when they threw away the old Netscape code base and started from scratch.

Consumers don't use an integrated Microsoft product until it reaches a point where they find it useful and interesting. Internet Explorer spent years underwater, and only started to rise to the surface with IE 3.0. They broke the surface and rocketed skyward with IE 4.0, a browser that most analysts concluded was head and shoulders above the competition.

The same applies to Media Player, which existed in Windows since the days of 16-bit Windows. Consumers mostly ignored it, and only started to pay attention when it became the complete music management program it is today.

I don't find many people in the TalkBacks who praise the RealPlayer client. Most appear to find it an inferior, even buggy, product. Think about what that means. Can you truly say with a straight face that the only reason people are ditching RealPlayer for Media Player is that Microsoft has hypnotised clients to download and use inferior products (Media Player 9.0 is not included with Windows XP)? Media Player simply is a better product today, and that fact has more to do with the shift to Media Player than the presumed advantages derived from integration.

Microsoft has lots of money, to be sure, and money can fund many forays into new markets. Smaller companies lacking such resources face unique challenges. However, the same applies in any market where small companies exist alongside larger ones. I would have a hard time competing with Dell from my garage. Yet, small companies still manage to grow into big companies on a regular basis.

I tend to think of the difficulties faced by smaller companies as akin to seedlings in a forest. We wouldn't hack down full-grown sequoias so seedlings can have an easier climate within which to grow (There is trouble...in the forest...).

Google, with a billion dollars in revenues, is hardly a garage-level organisation (or a seedling). Microsoft will apply the same technical effort it has to past contests, and build technology that will attempt to move beyond that which is offered by Google. If they succeed, then that is GOOD for consumers. Don't presume that consumers, however, will use Microsoft's offering just because it comes from Microsoft.

To Microsoft: Play smart
Microsoft today must operate in an environment where governments scrutinise its every move. Though the settlement between the Department of Justice and Microsoft declared that they have the right to design their own products, the courts have concluded that Microsoft's position, however fairly derived (even Judge Jackson accepted this), gives them market power. To mitigate that power, the courts declared that Microsoft needs to open protocols that are critical to interoperating with a standard chosen by 95 percent of the computer-buying public.

Microsoft claims they have made adherence to the settlement job number one. The upcoming contest with Google gives Microsoft a golden opportunity to prove it through the manner in which they choose to compete.

Microsoft ought to have the right to ship their search-oriented products with Windows. Microsoft spends ALL the money developing and marketing the operating system, and as long as they are a company in a theoretically free market, they get to decide what goes into their products.

However, the manner in which search functionality is reused throughout Windows ought to be open. For instance, one likely area of integration is to enable Internet searches from within File Explorer. I say: bueno...but design the integration in such a way that any third party can plug into that search mechanism and act as a provider for search data. Furthermore, allow users to configure from a central location what provider ought to be used. The default might be Microsoft's new search tools, but let consumers choose to select Google as their search provider (provided, of course, they've downloaded the Google provider, which as noted previously, IS NOT THAT HARD, and was done by millions with Google's search bar).

Microsoft is likely to integrate search functionality into its entire product line. Again, I say, bien...but make sure the same "pluggability" described in the previous paragraph applies in integration of search with the Office suite. If a Google "search provider" is selected as the default, make sure it is used in EVERY case where search functionality is requested by a product, Microsoft's or otherwise.

Popularise the fact that integrated search is a pluggable feature, and that anyone can plug themselves into the framework provided they follow certain interface conventions. This is necessary because many areas where Microsoft has built "plug-ability" get overlooked. As I noted in a previous article, Media Player is already a pluggable framework. There is nothing stopping QuickTime and RealPlayer files from playing within Media Player, besides a willingness to implement the interface conventions required to "plug in" to the DirectShow framework (the media handling API upon which Media Player is built).

Microsoft has always built every piece of software as operating system components so that third-party software developers might easily reuse them. They have also made most of their products highly extensible in order to enable developers to build atop them. Microsoft should continue this practice by being completely open and extensible in the manner in which search functionality gets integrated into Windows.

If Microsoft does this and is public about it, they might find antitrust regulators stop paying them so much attention, even if they still have a dominant market share.

Microsoft's competitive track record is often misrepresented. In some quarters, the very notion that Microsoft can do anything right is treated as blasphemy.

Microsoft, however, DOES serve an important role in the industry. As noted, it competes in more markets than it dominates, and furthermore, the nature of that competition is marked more by technical advances than simple distribution through its dominant Windows product.

That doesn't mean, however, that Microsoft won't avoid the functional equivalent of handing its critics boxes of ammunition. Microsoft has the advantage of size, money and programmers, and is fully capable of competing on pure implementation skill and features. They should trust that, and avoid the kind of actions that would provide justification to future investigations.

John Carroll is a software engineer now living in Geneva, Switserland. He specialises in the design and development of distributed systems using Java and .Net. He is also the founder of Turtleneck Software.