SEO Experiment

We have been reading about how SEO is changing to catch sites that are over optimizing using methods such as buying cheap irrelevant links, obtaining an advantage by weight of links. An interesting study in Webmaster World, looks at the traditional SEO techniques and assesses its results.

In summary, two websites were created one was developed aggressively using traditional SEO techniques.  The second site was developed using techniques which we might consider more  natural methods of building natural links  such press releases and social media. The SEO methodology for both sites is shown below:

Site 1:

• Brand new domain name that was brandable e.g. “widgets4you.co.uk” (the 4u and RUs thing is very popular in the UK!).
• We built around 70 doorway pages i.e. Pages on topics related to the widget being sold and wrote keyword full (not stuffed) articles.
• Paid links from directories (e.g Yahoo) and other high PR sites.
• We built a reciprocal (not three way) link directory on site offering link exchanges to anyone in any industry.
• We hired 2 people and had them work full time finding reciprocal link exchanges on other people’s sites that were mostly related to us … though sometimes the connection was tenuous!
• We wrote 10 articles and submitted them to article sites.
• Anchor text was therefore controlled and total links as of last month was 7,151.

Site 2:

• Bought a 3-year aged domain name on Sedo. The domain was a keyphrase i.e. bluemetalwidgets.co.uk
• We started off adding around 3 pages of unique and informative content a day aimed at our market, any keywords in there were all naturally occurring, not aimed for.
• After the first 6 months we reduced this to 1-2 pages per day. Again, each article was on current affairs, tips etc. for the industry, totally unique and between 400-800 words.
• We developed press releases and submitted a total of 16 over the last 6 months through PRWeb, PRNewswire and BusinessWire.
• We developed a blog.
• We developed a video news channel now showing 22 videos.
• We started aTwitter service of main headlines.
• We started an RSS/email subscription service to our articles.
• We had no link building work, we relied on natural link occurrence (4,374 links as of yesterday) and natural anchor text.

While there are many arguments about the validity of the site, the experiment demonstrates that in this example, that old techniques were still working and newer social media do not work. However that is too simplistic.

Obvious objections were that the previous domain history for site 2 was unknown.

While links were created they were from directories (e.g Yahoo) and other high PR sites. While paid links may not be to Google’s taste, they are still links and if chosen well should not be a problem. They had a good PR.

Links to any industry – can not say much about this. On the face of it, bad practice but it depends on who they linked to and how many there were? It is very difficult for a search engine to tell which industries should be linked or related.The quantity of these links is important and enough poor links could create a problem.

Same for the link building. Two people creating links in relevant industries, even if tenuous is still okay. Poor links may add some noise but not surprising.The quality of link building depends on the people that are creating the links and their understanding of the process.

All SEO tactics can be used well or poorly. A surprising result would have been if the site had been created from duplicate content on a brand new domain, linked to lots of spammy links to irrelevant sites with no PR, poor quality directories and then rank for keywords.

On the basis of one experiment it is difficult to draw solid  conclusions. But this experiment leads me to conclude not that there is a division between old and new techniques, rather that SEO if carried out well can have an enormous benefit.The experiment pits traditional  SEO against the social media techniques,  however they are rarely used in isolation. It would have been interesting to see the result from a site with a combination of both techniques.

Full discussion can be found at:

http://www.webmasterworld.com/google/3889779.htm