The dark arts of SEO and why Global Redirect won't help you

Every now and then a post comes up on Drupal Planet, or somewhere else in the Drupal community describing how the Global Redirect module (GR) is the key to improving your SEO (Search Engine Optimization) and how it will increase your search engine rankings. But don't be swayed by the SEO voodoo, most sites will see little to no benefit from GR.

Don't get me wrong, the basic premise behind GR is sound. Let's start with some basic search engine theory. If there are two pages that are theoretically the same in every sense of their search engine friendliness, except for the fact that one of the pages has more links pointing to it from other people's sites, then it will have a higher search engine ranking. It's difficult to describe so lets take a look at another example. If there are 10 links on other people's sites pointing to your site at URL #1, and 10 other links pointing to URL #2, then those two URLs will get lower search rankings than 20 links pointing to just one URL. So if those two URLs are in fact the same page, you could improve the situation by having just one URL. The problem is that there's a lot of misunderstanding about how this can happen in the real world.

Let me back up for a minute. SEO is a dark art*. The algorithms that search engines use to determine who gets listed for what searches in what order is extremely complex. Nor do they simply let everyone see how it works. Search is after all a billion dollar industry. Furthermore the traditional marketing wisdom of "make small changes and measure the results" is pretty much useless when it comes to SEO. Unless it's a really really big change, it's almost impossible to see whether that change made any difference in your rankings; There's just too many variables involved. And so there's been a load of misinformation about SEO on the web over the years. Unfortunately there's been a few bad apples in the SEO industry who have manipulated these vague facts to their advantage.

Back to GR. Let me talk about some of the problems that GR will fix, that aren't actual problems for most sites:

#1 a node is available at multiple URLs

Yup it sure is. This very page is available at:

http://advomatic.com/node/1002
http://advomatic.com?q=node/1002
http://advomatic.com/dark_arts_of_seo_why_global_redirect_wont_help_you
http://advomatic.com?q=/dark_arts_of_seo_why_global_redirect_wont_help_you

But unless you do something really strange like what I just did here, no one will know about these other URLs and no one will link to your site with them. They won't exist in the wild, so the search engines won't find them.

There is one exception and that is if your site has been up and running for a while either with clean URLs off, or Pathauto off, and you then turn one or both on. This is one case where GR may help you.

#2 Your front page has two URLs

Some sites have a problem where they have links to both

http://example.com/
http://example.com/front

or something of the like. Sure GR will make this go away, but that's really hiding the symptom, not fixing the problem.

Whenever you are making a link to the front-page either in code, in a menu item, in Views, or anywhere else in the administration interface you should use <front> instead of the actual Drupal path. that way Drupal will spit out the correct URL sans path. And, just like in #1, if that other URL isn't around, no one will know to link to it.

#3 If content is available at more than one URL the search engines will penalize you

This one is an example of the misinformation that has been floating around the net for a while. Google in particular has tried hard to make clear that this isn't the case, but every now and then someone still tries to state this as fact. Let me make it clear. If a page is available at more than one URL you might be able to improve its ranking somewhat, but you won't be penalized for it.

#4 GR removes trailing slashes from URLs

Umm, think about it, when was the last time you created a link, but then decided to put an extra "/" on the end just for kicks. Sure GR will catch this scenario, but how often is it going to happen?

#5 "But I have pages with more than one alias"

Ok, sure. For all sorts of reasons we make changes to our information architecture and things get moved around, but we still want the old URL available. I would suggest Path Redirect is a better choice. If performance is a concern a redirect rule in your .htaccess file is better, in your Apache config files is better yet.

"But what if my content editors change the URL when they edit the content?" This begs the question why do your content editors need control over the path? A copy writer or editor is not going to be an SEO specialist, nor should we ask them to be.

"I fall under an edge case"

Yes there are edge cases, some of which I've talked about above, where you may want to investigate GR. But when it comes down to it, SEO is a very complex subject and simply unifying a few URLs probably won't do much for you. If you're really concerned about SEO I suggest reading a blog series by our friends over at Civic Actions.

Furthermore there are other things to consider like performance - more code is slower, and maintainability - more modules means a more complicated upgrade path (maintainability is probably a subject due for it's own blog post).

Most sites will see little see little benefit to GR, and most sites hold performance and maintainability as much more important than SEO. And because of that I have yet to see a case where GR would be useful for any of our clients.

SEO is a subject shrouded in mystery. What are your thoughts?

*the term "dark arts" is stolen from Marco Carbone

Michelle wrote 5 years 28 weeks ago

#4 actually happens quite frequently, though not in the way you wrote. Pretty much every time I paste a link into a forum post or a blog or a tweet is has a slash on the end. Why? Because I copy it out of the address bar.

Michelle

Dave Hansen-Lange wrote 5 years 25 weeks ago

I'd be interested to know what browser you're using, if it does that's a big bug because
http://example.com/some_file
is not the same as
http://example.com/some_directory/

However Apache will by default change
http://example.com
to
http://example.com/

But it does so in a way that everyone gets redirected to the second URL and the search engine will only index the second URL regardless of which form people enter in links.

Jim wrote 5 years 28 weeks ago

Added to DrupalSightings.com

greggles wrote 5 years 28 weeks ago

For #3 - Are you _sure_? How do you know? I've heard both ways on this from various people that I respect - but is it worth taking the risk? Being a position or two higher in the SERP can mean 3 times more traffic.

I also think you come to a bad conclusion because this article only considers SEO. It is also beneficial for regular users of a site to only have content available at a single URL.

Global redirect also provides a potential benefit for building a "tinyurl" like system in Drupal. I often send links in e-mails as http://growvs.com/node/243 so that they don't wrap across multiple lines.

When considering the downsides, you mention performance and maintenance. I'm not aware of any real maintenance issues with Global Redirect other than having to ugprade "one more module" when going from 5.x to 6.x. So, I think the maintenance issue doesn't carry much weight.

For performance, have you benchmarked global redirect? In my experience it's never been the source of a slowdown.

I recommend always installing it because with no major downsides it may help SEO, it can also help a site's usability, and it can be a nice benefit for shortening urls in mail and twitter.

Dave Hansen-Lange wrote 5 years 25 weeks ago

Agreed there is a lot of varying information on the topic of search engines penalizing you for dual content (or not), so it's possible that I'm wrong. However I would find it hard to believe that it would make a significant difference on the SERP. Content is king.

About URLs in emails yes that may be a case where it may help you if a lot of people are copying URLs from the emails to a page on the Internet that search engines are indexing. But realistically how often does that happen?

My main point with this article was that many people don't really understand SEO. There's a lot of false claims out there that tools like GR will greatly boost your SEO by fixing problems that, while they may exist in some rare edge cases, in reality aren't really a problem.

Spending 5 minutes reworking the copy on your homepage might have a much bigger impact on your SEO than these little technical tweaks.

Jeff Geerling wrote 5 years 28 weeks ago

Especially with regard to #4:

A lot of people (myself included) *do* type in URLs with ending slashes from time to time. If someone creates a link like that, then it will redirect to a slash-less URL, and then if someone else copies that URL out to digg the posting, or create a link from their own site, the link will be uniform with links that other people see when they visit the site.

Global Redirect simply tidies up the 3 or 4 little things that aren't absolutely necessary for SEO, but are like icing on the cake.

The most important part of SEO, of course, is content—well written content, with good headings and page titles, beats out any plethora of SEO-optimization tips, modules, guidelines, etc.

yaph wrote 5 years 28 weeks ago

It is not correct that no one will know about these other URLs unless you do something strange . If you look at your server logs you will see requests to URLs like node/123 even if no single link on your site points to these links.
Search engine and other robots gather information about your site by requesting URLs that not necessarily need to be linked.
Furthermore, can you be certain that none of the modules you install on your clients sites' links to node/123? The twitter module for example posts node/123 links to twitter. Twitter links are nofollowed but that does not mean that robots do not follow them.
The global redirect module takes care for you in these cases, whether it is a module that does something strange or your editors and it comes with very little processing overhead.
I think if you use path / pathauto you should use global redirect too. Search engines may not penalize you because of duplicate content, but it certainly does not help to have duplicate content.
Do you know where algorithms draw the line between what is considered intended duplicate content and what not?

Dave Hansen-Lange wrote 5 years 25 weeks ago

I think you may be confusing legitimate search engine bots with other bots. The bots of all the major search engines play by the rules.

Neil wrote 5 years 28 weeks ago

You forgot:-

http://www.advomatic.com/comment/reply/489

Contact Us

About Dave Hansen-Lange

Dave Hansen-Lange has been developing websites since 2003 when he needed a web presence for the record label that he founded with several fellow musicians.

AdvoTwitter

  • RT @wiredimpact: Tips on choosing stock photos for your #nonprofit website that aren't cheesy http://t.co/cpFYWRIicG #nptech
    October 1, 2014 - 1:49pm
  • Looking for a career in the progressive movement? Check out @NewOrganizing Institute's NYC Career Fair! http://t.co/d8DwkVnmbd
    October 1, 2014 - 10:15am
  • #DrupalCon Amsterdam is happening right now, and we're super bummed we're not there. Just gonna follow the hashtag and think happy Drupal...
    September 29, 2014 - 1:16pm
  • NASA's on #Drupal now! http://t.co/v4NoMoLHlj
    September 26, 2014 - 3:15pm
  • Join the Women in #NPTech community call today at 3pm ET! http://t.co/V25NyyeMeM
    September 25, 2014 - 10:15am