There was a time when measuring SEO performance was no big deal. Marketers just collected the top 20, 40, or 100 high volume keywords and focused on those to improve their SEO performance. Then about three years back, things started changing. Looking back at the changes that Google made, it is now obvious that Google wanted to strongly discourage any actionable measurement of SEO. With that in mind, they made it really, really hard to analyze SEO performance at a tactical level.
So it is time you upped your game as well. There is an air battle raging. So drop your swords and upgrade your arsenal. This is a four part series on how to get smart with SEO analytics. In this post, I describe the big changes that have stymied SEO analytics, and then I touch on the three three key ways you can upgrade your organic search channel's analytics. So let's get into how Google made SEO analytics hard.
Google Changes That Impacted SEO Analytics
Here are five key Google changes that have made SEO performance measurement really hard:
1. Google took away tracking at the keyword level altogether. Now they hide information about keywords searched by clumping all that data in Google’s (not provided) bucket.
2. Google introduced Auto Suggest to encourage people to do more long tail (three and four word search phrases) searches. While the reason here was to make search results more relevant, it took SEO analytics beyond the scope of simply using the top 20 keywords to gauge performance. Now performance needs to be measured after taking into account the many combinations of words that people might search for.
3. And then they implemented semantic search. That’s just a fancy word to say Google now understands the meaning of a word within the context of a search. For example, a search for “fix air conditioner” shows results that have “air conditioner repair” in the content and not the word “fix.” So Google now understands when to use alternate words to display the most relevant results.
4. Google also made search results geographically relevant. So I get a different set of search results when I search for “carpet cleaning” in Dallas versus someone searching for the same service in Chicago. This means the generic national level results that are used to determine performance are not actually relevant to what’s actually happening in Google search.
5. Google personalizes search results based on search history. This is probably the hardest change to work around because each user’s search result could vary wildly. In fact, if you look at the keyword ranks in Webmaster Tools, you’ll see some weird ranking because of this very reason.
And the list doesn’t stop there. All these changes have made SEO analytics for even small websites fairly complicated.
Now there are marketers who say that one shouldn’t even try to measure SEO at the keyword level. They haven’t spoken to your CEO, have they?
Seriously though, three distinct SEO analytics artifacts are needed to paint a full picture of SEO performance today. They are SEO Demand Analytics, SEO Conversion Analytics, and SEO Link Building Analytics.
SEO Demand Analytics
This is the practice of understanding how people search for products or services your company offers and where your website stands compared to your competition. With Google not providing keyword level data anymore, marketers have turned to other various sources to get a feel for the types of search terms people use when searching. Some examples of such sources are SEMrush and SERPs.com. Solving the data source is just a small part of the problem though. Analyzing keywords that include location modifiers (e.g. carpet cleaning Dallas) and long tail phrases (e.g. commercial carpet cleaning companies Houston) take analysis into large data territory.
I will cover how to run SEO Demand Analytics for large keyword data sets (from 1000 to 1 million+) in an upcoming post.
SEO Conversion Analytics
Simply put, understanding how users convert once they enter the site from the SEO channel is what SEO conversion analytics is all about. But it is never that easy, is it? Here are some of the complications when trying to analyze the SEO channel’s conversion:
1. Brand traffic behaves differently from non-brand traffic. Brand traffic is likely to convert better than non-brand. However, now that Google doesn’t provide keywords used, there isn’t a way to directly differentiate brand traffic from non-brand traffic anymore. So it becomes hard to figure out how non-brand SEO is actually performing.
2. When the same page ranks for a range of head, mid, and long tail keywords, its conversion rate will vary across this large set of keywords. At that time, it becomes very hard to separate the high vs. low converting keywords.
3. SEO is a huge driver of content seeking traffic. This traffic may not necessarily convert the first time, but it might in subsequent visits. This is particularly important to measure in SEO because conversion “assists” need to be attributed back to SEO.
As you can see, just selecting the medium as “organic search” in a Google Analytics segment doesn’t really provide us with a reliable picture of SEO conversion. More about this in a future post.
SEO Link Analytics
While it is very important to have good, fresh content on a site to make it Google worthy, the real secret to ranking well is to get high quality links coming back to your site. So it becomes very important to understand where exactly your site stands in comparison to the competition. Fortunately, Majestic SEO provides a wealth of information about link quality to give you an idea of how your site stacks up.
So what makes SEO link analytics so complicated? Plenty of things:
1. Given that the name of the game is popularity, link analytics has to be done by comparing links of other sites to your site’s links. The problem is that for even a “medium” sized website, you’re going to need to sift through thousands of links to determine which ones are worth the squeeze.
2. Time plays a critical role in the link building world. Analyzing another site’s links to understand their strategy doesn’t always work because Google moves fast to patch up holes. So what worked six months ago may not work today. In addition, the discovery of the link by Majestic SEO may not be at the same time that Google discovered the link.
3. It is a lot of manual work. What’s manual is visiting the linked domains and understanding whether the domains are worth linking from.
I’ll get into how to do SEO link analytics in a later post.
A Last Word
To recap, these are the three critical aspects of a solid SEO Analytics practice: SEO Demand Analytics, SEO Conversion Analytics, and SEO Link Analytics. The problem for many marketers is that they don’t have the necessary skill set to set up analytics to get a comprehensive view of SEO performance and quickly identify what needs to be worked. Indeed, many still use the top 30 keyword approach today! In the next few posts, I will walk you through how to manage and create the three artifacts of SEO Analytics for your site.