We've been out of private beta for only two days, and stackoverflow.com is already ranking high in Google Search.
Are there any specific techniques applied in the site? I've been researching the subject for some time and haven't found it to be so easy.
Searching 'stackoverflow' even returns site-links.
Disclaimer: I find this a very interesting topic, so I've scanned through every single answer, and now I want to summarize here what I've just learned and group all good links in one place.
And make no mistake, Search Engine Optimization today is all about Google, even while people want to avoid saying it (just count words for "Yahoo" or anything else there).
"The Golden Rule of SEO: Content is King" is the first thing to consider. It's quite simple: have good content. That's the primary reason Stack Overflow has a chance to rank at Google, simply because of how page ranking works.
Websites that have no optimization or guidelines concerns whatsoever but got nice stuff have a great chance of ranking high, because there will be people linking to good content. And most sites trying to do SEO without content won't rank well because nobody will link to them.
With that been said, Stack Overflow does a lot of work from a long time ago to keep clean HTML code...
One good example is how rel='canonical' is applied in here to keep search engines away from linking to duplicate content. For instance, let's get this page source:
<link rel="canonical" href="https://meta.stackexchange.com/questions/14056/seo-in-stackoverflow">
That is an optimization that will not help to directly rank in the search engine, but it will make the results the engine can find much more relevant and through that it might generate more links (from people who found what they were looking for) to the good content it already have, thus actually indirectly improving its rank.
So clean and optimized HTML does work in conjunction with good content...
But content comes first.
SEO people say "Content is King" - SO is all about the content, which grows and changes by the minute. This is also something that ranks high with Google, and as Daok points out, having lots of high profile sites linking to it, Google rates it highly.
The single most important factor in a site's ranking on Google is how many sites link to it. Stack Overflow is being discussed relatively heavily in the tech-press and blogs, quickly giving it a high ranking.
See: The Golden Rule of SEO: Content is King
Stack Overflow already has a lot of great content. Also, incredible niche topics. See Long tail (Wikpedia).
I don't think Jeff is doing anything magical here to make this happen. It's just good site design and popularity:
Overall, the site is just well designed for SEO.
Having an XML sitemap (which this site does) helps immensely, because it helps the Googlebot find dynamically generated content much more quickly than it would otherwise. You can see stackoverflow's sitemap here: https://stackoverflow.com/sitemap.xml but be prepared as it is a huge file.
RSS feeds also help a bit, as Google pulls them as part of the blog search engine. I recommend all sites with content that updates regularly have RSS feeds, even if you are not even close to being a blog.
One thing I noticed right away that's supposed to be good SEO is putting the question name in the URL. You can get to the question using just the question number (for example, this post can be accessed at https://stackoverflow.com/questions/86268/) but the default URL has the post name in it.
Google TechTalks recently hosted a presentation by Joel Spolsky where he talks about stuff like this. Worth watching: Learning from StackOverflow.com
[Edit - Watch the whole thing but the question is addressed in the video starting at 23:45.]
See Jeff Atwood's post on his blog, The Importance of Sitemaps. I also noticed that Stack Overflow questions are usually indexed by Google within an hour. Try googling for the title of a recent question.
Stack Overflow has a growing community and more and more people talk and write about it. All those links and users bring importance to Google (and the PageRank of Google).
Most companies employ "blackhat" SEO. Gaming the system. Employ "whitehat" SEO. Make a great site, make content that people want, design it so they can access it well, and good rankings will come naturally.
See High Accessibility Is Effective Search Engine Optimization (published 2005-11-08).
In one of the blog posts, Jeff mentioned about The Importance of Sitemaps
While Jeff A. says he doesn't do 'SEO optimization', stackoverflow does follow about every search engine recommendation, and there has clearly been some effort to make the content indexable. For example, every page has a reasonable title and URL. Typical web frameworks will create urls like /question/12345. On SO, they create URLs like Digg. They also use the rel='canonical' link to keep search engines from created duplicate content.
Sitemaps helps, but the reality is that Google ranks sites both by how frequently they update as well as the value of the content (PageRank).
Sites with high pagerank that update frequently (such as CNN, and this site) get crawled more frequently, and in some cases continuously.
Site maps makes that process even faster/easier for Google, and so that content goes up that much more quickly.
This is very, very anecdotal; however, usually when I search for something programming related that is easily expressed or is very basic question, Stack Overflow is on the first page if not in the first three links.
Quite often I do not go to Google first anymore, but rather to Stack Overflow. If there's nothing or not enough relevant stuff here, then I go to Google.
If there's not much on Google, I might ask here and continue searching Google while waiting for hints. It seems to be working well for me most of the time.
I'm supposing that the site was submitted for indexing at the very least, but I'm sure that with the traffic SO is getting, Google incorporates that data into its algorithms.
Just thinking about how many people probably have some Google related piece of software on their system, I'm sure it gathers some sort of (hopefully!) anonymous data about sites visited to bump a popular site up on the indexing ladder.
I'm not sure that those examples demonstrate that Stack Overflow has questions actually being ranked high by Google yet. The first search is directly for Stack Overflow, which naturally ranks it high. The second is for a particular misspelling of database: "Dabtabase" which happpens to appear on a Stack Overflow page.
SO marks most outbound links rel=nofollow. This would inflate the sites ranking according to PageRank (assuming out links marked nofollow count as not having an out link at all), if Google made no compensating adjustment to it (say, 90% of out links are rel=nofollow, which most quality sites don't do, hence..)
