German interview on Google

I did an interview about Google for netzeitung.de, a German news web site. If you don’t read German, I’ve included an English version of what I sent them:

Q: Mr. Kottke, how far away is Google from “being evil” in general in your opinion?

Google is doing better in the corporate morality department than a lot of other companies. From all accounts, their leadership wants to not be evil, they treat their employees well, they take great pride in the usefulness and relevancy of the results of their free search service. The terms of service for their AdSense program is definitely a step in the wrong direction, like they’re letting marketing and legal determine their approach to business instead of the other way around. But Google is a long way from a Verisign/De Beers/Enron level of evil.

Q: But that was Google’s rule number one, wasnt it?

I think every corporation’s real #1 rule is “make money”. That Google wants to make not being evil an equally important priority is commendable.

Q: How effective is Google still, given the springing up of Google Spam (lots of doorway pages leading to sponsored links) and the “noise” coming from weblogs etc.?

Google’s search results are at least as relevant as they have always been, if not moreso. I can almost always find what I’m looking for in the first 10 results or so. I think part of the problem is perception. The perception is that their results should continue to improve as they refine their search methods and algorithms, but there’s an inherent limitation in their approach that limits the maximum possible utility. There’s only so much information about how pages relate to each other that you can glean from scraping web pages, and if Google is close to reaching that limit, any changes they make will only result in small changes in usefulness.

Q: Can Google do anything about this?

Perhaps they might want to start grouping web pages and sites into groups and analyzing the sites in each group in a unique way to improve the overall database. Weblogs are a good example of a group that could be analyzed differently. Weblogs consist of separate posts, which should be treated individually to get the best possible data from them. Weblog posts often contain more metadata than a typical web page, things like date and time of publication, categories, backlinks, etc. Google can use that post-level metadata to get better information about the sites that weblogs point to without having to pump the weblogs up in the overall rankings — as many people have complained is not so good. Many weblogs also have RSS feeds with structured metadata that could be analyzed to improve general search results.

Q: Does Google need more competition?

A little competition for Google would be a good idea. Microsoft and Yahoo have both announced efforts to improve their search engines, but I don’t see them developing anything to threaten Google’s search.

Q: What will the Google of the future look like?

Given that the look of their site hasn’t changed significantly since the beta version, I wouldn’t look for it to change much in the near future. The biggest change will probably be more personalized search results where my results for a given search would be different than yours based upon our usage of the site.

[Hopefully that all makes more sense in German.]