The cost of changing URLs

Every week people look in SEO forums for help with decisions on whether to change URLs for contetn on their sites. Some people want to change domains completely and they ask if they’ll lose their rankings.

Well, we know you lose your rankings, so the real question comes down to: what’s the return on your investment? What is your business case for rebranding?

I chose the Blogger service for the original SEO Theory blog for several reasons. First, everything published on Blogger shows up in Google Blogsearch in less than a minute and I’ve seen it hit Ask’s blogsearch in about an hour. Secondly, services like Blogger and Wordpress send you random traffic. Traffic is good. Third, it’s easier (in my opinion) to get the search engines to pay attention to sub-domains that are hung off of very strong, well-indexed domains.

But when Google upgraded Blogger, they pretty much ruined the service. The cost of losing crawlability on a service needs to be crawled continuously was just too great. So here I am rebranding the SEO Theory blog, having to create a Technorati Profile so that I can claim ownership of the blog so that Technorati will index the site within a reasonable period of time.

To me, a blog should be indexed in an hour or less. I prefer a minute or less. Sorry, but Blogger spoiled me. I could generally expect my Blogger posts to show up in Google Web search in about 3 days. So moving to a new domain has cost me visibility, and I knew when I made the move that I would have to give up that visibility and traffic.

Still, here I am, building value in yet another domain. I can think of other things I’d rather do with my time. It’s not like I need the practice.

Fortunately, I’m not selling anything off this blog. I don’t have to worry about lost revenue. But I understand why people agonize over the decision to rebrand content. It’s a time-consuming process and until you’ve gone through it you really have no idea of what it will cost you.

In about a month, give or take, everything should be okay for me again. I’ve already redirected many links from the old SEO Theory URL to this new one. And the SEO Theory pages at Blogger have meta refreshes and robots tags telling the crawlings to come look here. It’s crude, but at least I don’t have to mess with .htaccess files.

I recently rebranded about 30,000 pages of content on one of my personal domains. Instead of moving those pages to a new domain I put them into a new set of sub-directories. I did this because the pages were way out-of-date and they lacked the type of strong internal linkage I have so often advocated that people should be using on their sites. To ensure I did not completely obliterate content that my visitors want to find, I set up a parallel content situation.

In fact, whenever I rebrand I go with parallel content. I use robots.txt and/or other means to tell the spiders to stop crawling the old content but I don’t remove that content until its traffic has nearly dropped off completely. I build important links from root pages and site maps to point to the new content as soon as it’s up. I remove all internal linkage (still being crawled) to the old content that I can get to (updating old forum posts is too much trouble).

For a brief period I may actually see a spike in combined traffic as the new pages are crawled, indexed, and listed alongside the older pages in search results. But eventually the search engines stop showing the older URLs that they have been blocked from and the new content stands on its own.

The one most common mistake people make when they rebrand is that they immediately take the old content down. Leaving that content up — even using it to point people to new content — reduces your lost traffic considerably.

Another common mistake people make when rebranding content is they try to pass on all their link value from the old URLs. The best approach is to create link visibility for the new URLs. Odds are pretty good that your new anchor text will be more relevant to whatever you’re saying now than your old anchor text. And if you operate two sets of data in tandem, the old inbound links are still useful in that they send traffic to pages you control.

Many people also often overlook the value of their Error 404 handling. Through the years I have experimented with different approaches. Presently, I’m using custom 404 pages on my personal sites that provide links to the site maps and popular destinations in the hope that people will cilck through to find the content they are looking for. Confusion over correct URLs is a common problem for people who are navigating through rebranded content.

I still have more work to do with this blog. I have to figure out how I want to track the statistics, for example. I may stay with Google Analytics or I may rely on raw server stats. Technically I could do both but I want to think about my options.

In the meantime, there are many links pointing to the old blog address that will probably never be updated. That is what rebranding has cost me. I’ve had to start over from scratch. Regardles of whether you use 301 redirects or any other kind of redirect, people will remember the old URL that you built brand value for.

It takes time to train people to instinctively associate a specific URL with a Web site. When you think about the costs of rebranding, be sure to budget enough time for the re-education process. That will be your highest cost of changing URLs.