Did you know that there are a handful of simple reports you can check in Google Analytics to see if your search engine optimization guru (or Agency) is catching critical pieces of the SEO puzzle? Take quick a look to sniff out what might be missing in your SEO effort.
Uncovering technical SEO issues
Of course, the wise SEL audience needs no reminder of the importance of the technical side of SEO. If the engines can’t easily see your content and associate discrete content with discrete URLs, mama ain’t happy. Fortunately for us, there are reports in Google Analytics that can also be used to help you uncover these technical problems, and we’ll talk about a few of those today.
How to identify canonical issues. There is a report in Google Analytics called “Hostnames.” It’s found under the Visitors > Network Properties section. If you’re having trouble finding it, here’s a picture of where to find the Hostnames report.
What we’re looking for here is a very short list of names that represent the domains we always intended people to see. Here’s a screenshot from my blog; please bear with the shameless use of my own sites—my clients aren’t too keen on me showing off their information.
We like what we see here. Almost. I decided to get cute with the name in some links and capitalize it, so we see two domains here where we really should see just one. Big deal? Maybe not, but I’ll leave it to the pro SEOs to comment. Should I fix it? Probably.
Now, let’s look over at another site—this one’s up on cinder blocks and there’s a little trash in the yard:
This screenshot reveals a big problem: a canonical issue with a www site and a non-www site. I need to 301 redirect traffic requesting http://evanlapointe.com to http://www.evanlapointe.com (or the other way around—just pick one). The IP addresses we see in there aren’t an issue—they’re usually just Google’s IP addresses from when people view cached copies of the pages. If you see a lot in your own results, just take a look at them to make sure. That last result? Not good, either. I’m going to have to give my hosting service a call.
Another thing you might see here are staging servers (like dev.site.com) or mirrors (like server2.site.com). Double check to see if you can access those from the internet directly. If you can, use robots.txt or the appropriate meta tags to disallow the indexing of those pages while you figure out a way to put them behind some sort of protection—not only do the search engines like to take a gander at all of your content, your competitors might find out about your new features before you do if they’re able to check out your development site.
Dig up some duplicate content. Another technical SEO enemy is a single piece of content with more than one home. This can be a particularly big issue for sites publishing articles, displaying products using parameter-driven pages, or if your site is just a quilt of many developers’ efforts. While Google Analytics doesn’t make finding these issues as simple as identifying some other problems, here’s a quick routine I go through with our clients.
Under the Content section of Google Analytics, I’ll pick a dozen or so pages under the Content by Title report. This report gives me the same information as the Top Content report, but it aggregates against the title tag instead of URLs. Again, likely candidates are articles, products, category pages or anything a CMS generates dynamically.
When you click on a title in this report, you’ll see the URIs (URIs are the domain-less URLs seen in Google Analytics) this title tag represents. See more than one page? Not good. See more than 3 pages? Not only do you have potential duplicate content, you may be missing one of the most fundamental SEO tactics: picking unique and descriptive title tags for every page on your site. If this is you, do not pass Go, do not collect $200. You’re already on the right site. Read. A lot.
But if you only see 2 or 3 URIs here, it’s likely that the same page (content) has several homes (URLs). Often enough, SEOs will create “friendly” URLs (like http://www.site.com/sunglasses/polarized.html), and forget to 301 redirect the old URLs (ugly ones like http://www.site.com/products/category018/product.jsp?prodID=2265) to the new ones. Then the engines come across a link to the old URL somewhere on the internet and there you have it: duplicate content. Make sure to redirect the duplicate URLs to the canonical (“authority”) URL, and make sure you have unique title tags for each unique page of content.
Beware: SEO can skew analytics reports
The best SEOs realize that analytics reports may not always be accurate, but many do not: all of this redirecting can potentially cause some major integrity issues in Google Analytics, or any analytics tool for that matter. In some cases, redirects may strip off critical URL parameters (for tracking paid search or other marketing campaigns) or prevent the analytics tool from determining where a visitor came from (often reclassifying referral traffic as direct traffic), and this is obviously a huge problem for the SEO who is trying to prove the traffic lift that results from their efforts. If you are seeing decreases in traffic after redirects are put in place, run a few tests to make sure that these redirects are working properly.
I’ll take some time to cover more ways Google Analytics and other analytics tools can help with technical and linguistic SEO strategies in future posts, but like my best teachers, I’ll end with a good strike of the ruler across your knuckles. Don’t forget to tag your paid search campaigns so your analytics tool can tell the difference between paid and natural search. Google provides instructions on how to do this, as do the other tools you may be using, and not doing this means that you’re just going to be looking at search soup, making it much harder to prove the value of either program.
In the meantime, if you have questions about doing the above in other tools like Omniture SiteCatalyst or working through these exercises, feel free to comment below or reach out to me on twitter.
Opinions expressed in the article are those of the guest author and not necessarily Search Engine Land.