post

This post should have a subtitle:  How I Broke My Website.

Even before Dr Jekyll self-dosed himself to become Mr Hyde, the idea of self-experimentation runs through our culture.  As any fan of comic books know, without self-experimentation, half of the super heroes of our fantasy world would never have come into existence.

In the world of search engine optimization, our industry has laid out a series of best practices, based on guidance from the search engine companies themselves and from the experience of practitioners.  And, like in the world of science, it is only through experimentation that we can really understand the world around ourselves, and improve our base of knowledge.

But, when someone is paying you to improve the returns of their website, it would be unethical to try experiments on their website just to see what will happen.  So, we need another way.

In my personal case, I have access to a second business.  I am co-owner (with my wife) of The Turret Bell.  This is a small seasonal store that specializes in Prince Edward Island Books and Music, and has a year-round on-line sales outlet.  For years now, I’ve been using this site as my own private little plaything, trying different SEO and analytics ideas.  Over time, partly as a result of my experiments and partly because of some fundamental problems with the structure of the site, it had become a bit disorganized and unfocused.  So, I embarked upon a complete redesign of the site.

And this is where I decided to play Dr Jekyll, and break some of the rules.  The only one who would be hurt would be me and my site, and in the end I’d have a better idea of what would happen if I wasn’t careful.

In this case, I wasn’t changing the location of the site.  It would still be at www.turretbell.com.  But, I thought I’d drop in the new site and replace the old site without taking basic SEO precautions like page redirects.

I tracked what would happen with a handful of keyword phrases, items that we previously ranked highly for, such as “Prince Edward Island Books”.  I fully expected to see a drop in ranking, but was a bit surprised by the impact of my experiment.

First, within a day or two, Google Webmaster Tools started screaming at me about broken links, since some of the old pages were still resolving, but the links off those pages were dead-ends.  Then the ranking dropped.  For instance, for “Prince Edward Island Books”, we were in the top two or three before the change.

After the change, we dropped right out of the top 50.  Essentially, we became invisible.

I was expecting a hit, but the extent of the drop took me by surprise.  It really emphasized to me the importance of planning and carefully executing a strategy for making major changes to a website.  This little experiment sure taught me a lesson about making sure that everything is lined up carefully when updating a website.

Unfortunately, it’ll now be a bit before this website recovers.  But I certainly have a better appreciation now for the potential consequences next time a client comes to me with plans for a new website.

In the meantime, if anyone wants to purchase some PEI Books and Music, I have a website that could use the help.