?

Log in

No account? Create an account
Semiformalishmaybe

Examining Google's success

Beyond the relatively uninteresting-but-important thing of being a search engine (I know there are some interesting parts but overall it is not flashy), I think Google's success at being more than a search engine comes down to three things:

  • Designing and adopting software institutions for effective, low-management, high-efficiency distributed code. This is sophisticated and interesting, but it's not obvious why someone would want it until one sees what Google has done with it. The failure of other software-as-service vendors to do this as generally hampers their flexibility and leads to high development, deployment, and maintenance costs for any service they decide to one-off. Google has published at least some of the code to manage this - even considering that they don't publish everything, the risk is low that they're helping their competitors too much because alongside the software infrastructure they've built hardware to its spec for long enough that they have a good lead.
  • Designing and adopting software institutions for a powerful user interface in the browser. This is also sophisticated and interesting, and also not obvious until one sees why someone would want it. Apart from the search engine itself and very high end applications (e.g. Google Earth), all high profile google applications rely on this and would be very difficult to imagine done without that toolkit.
  • Keeping a simple design æsthetic. Other search engines and internet services have come and gone, many of which made their sites prettier than google's. We should understand that as vanity - google's software is low-frill enough that it feels like a utility, with the very occasional quirks kept to a minimum and nondisruptive (e.g. custom google logos per certain days of year). Except when the user's network is bad, the Google experience is as fast and reliable as flicking a lightswitch
Any of Google's competitors could probably have done these things and been roughly as successful - the other specifics of Google are not necessarily significantly helpful (or helpful at all) towards their goals. I don't believe Google can live up to its 「don't be evil」 motto when their primary source of revenue is advertisements (adverts being intrinsically harmful). The tendency not to charge end-users for products helps them get mindshare, but if software is good enough at some useful tasks, it can still be expensive and thrive (e.g. Lotus Notes).

I recently reluctantly moved some spreadsheets for our lab on campus from GNUMeric on my laptop (which I utterly love) to Google Spreadsheets. The collaboration stuff is well done enough to make me forgive the interface.

Comments

During orientation, among the things that google had beat into us were the following: http://www.google.com/corporate/tenthings.html

There's a lot of wisdom in those, especially the first one. When considering a certain aspect of what google does, it's clear that it works so well because of all the thought and effort that's gone into making it as good for the user as possible.