Looking through all the analytics since moving www.alexanderchalkidis.com/blog old stuff to www.alexanderchalkidis.com/dotnet it was obvious that Google bots were not going to figure this out for themselves. First I put in a few links in other blogs to see what would hapen. Nothing. If you search for any older articles on Google, you get the old link. Even after a month! So I put more links in other articles, even really popular ones at http://alexartisia.wordpress.com and other blogs. Obviously the free version of WordPress behaves very differently to a properly hosted one, SEO optimised and all. So then my new www.alexanderchalkidis.com/blog , which is all properly WordPress hosted and optimized started featuring links to the new location. Still Google bots weren’t picking up.
So I shot off a Joomla website to test out how it behaves. Completely prematurely and it looked terrible at first . I did not update any indexes or submit anything to Google Webmaster tools and see how long it takes them. It was almost instant thanks to a few well placed links to older articles. Essentially, what the machine had to figure out is that any link to my old blog can be easily converted to the correct new location, simply by adding a www.alexanderchalkidis.com/dotnet at the beginning and replacing the “aspx” ending of the file with “html”.
So: http://www.alexanderchalkidis.com/blog/post/2009/12/Will-Xing%2c-Viadeo%2c-LinkedIn-or-Facebook-win-the-networking-war.aspx is now http://www.alexanderchalkidis.com/dotnet/www.alexanderchalkidis.com/blog/post/2009/12/Will-Xing%2c-Viadeo%2c-LinkedIn-or-Facebook-win-the-networking-war.html
In bold the bits you have to add or change to the old location to get a new – working – one. My six year old son can probably do this, Google bots can’t. Then again the whole point of the exercise is to increase targeted results without paying a penny in Google AdWords so maybe they don’t want to!
While in retail with Public I really got excited about the experimental approach to business. Set an experiment up, test it, adjust, measure, tweak and again. Properly done in retail it is phenomenally useful. Now I am using a similar techniques with SEO. The way I handled this change, total visits to www.alexanderchalkidis.com fell dramatically. From around 800 on an average day (peaks are 2500, lows are 450) it dropped to less than two hundred!
This gave me a unique opportunity to test assumptions about where the actual traffic is coming from.
1. Several permanent visitors which I thought were regular fans, turn out to be corporate (PR agencies probably) searchers, checking whether I have written something about them every day. From the looks of their queries, this is done automatically. Hey, that’s what you get for writing nasty (though true!) things about people!
2. My main loss is articles in minor blogs or websites which are not following up on their broken links, or not bothering to update them. (And just deleting them as they don’t work.) They were sending me a very healthy 30% of my traffic since several articles were deemed as “unique in their perspective”. These were articles I wrote specifically to examine how necessary a “other” opinion was in the cyber world and how it would circulate. Things like questioning whether eye laser surgery is really worth it which may have plenty criticism in the US but not in the Greek language.
3. Several other websites and journalists have tagged me by topic or category. I am obviously heavily plagiarised, thank you very much for the honour! Most do include a link to the original article. Now if only they would update it… Google searching for one of my articles is up to 70% of what it was before the switch and rising rapidly.