The Evolution of SEO Metrics by @sams_antics

If you invest in SEO in any way, shape or form, then you’re almost certainly using some sort of metric to track the success of your campaign. The real question is this: Which metrics are you tracking, and are they the right metrics to track?

Once Upon a Time

During the first decade or so of SEO being a thing (mid-90′s to mid-2000′s), it was pretty unusual for a site to have web analytics of any kind unless it was a major site. There weren’t many good options and the good options weren’t cheap.

Google Analytics didn’t come online until 2006, so from the dawn of the web to 2006, most site owners didn’t have any data available that they couldn’t pull from a log file (which often wasn’t the most helpful data, especially in the late 90′s and after).

As SEO started to become a lucrative business, with little or no analytics data available, these SEO’s needed to find a metric that they could track without analytics and that they could also directly influence. Enter rank tracking.

During this period, search results were pretty much all the same: 10 plain text links with basic descriptions. Because search engines used the same format, it was quite easy to track where a site was ranking (by scraping search results) and to then correlate that ranking position data with traffic, conversions, or whatever other data was being measured on or offline.

Because of the uniformity of the SERP’s, there was a very clear correlation between ranking position and the other metrics being tracked, and so rank tracking became the de facto SEO metric to measure, which was great then, but not so much now.

RIP Rank Tracking

Rank tracking was a great thing when search engines all used 10 blue links, but those days are LONG gone. In fact, in his recent presentation at MozCon, Dr. Pete Meyers gave a presentation titled Beyond 10 Blue Links that shows just how vastly search engine results have changed, and consequently, why rank tracking is no longer of value as a standalone metric.

Dr. Pete found that Google actually has 85 different SERP features above and beyond the de facto 10 blue links and that 85% of all searches turned up one or more of these additional features.

Because of the way these search features work and the impact they have on CTR’s, a #1 organic ranking position can actually appear below the fold of the page, dramatically reducing your organic CTR for that position.

Knowledge Graph in particular has had a tremendous diminishing impact on organic click-through rates, as aggregator sites everywhere can surely attest:

Now, these changes aren’t all new, though new things are popping up constantly. Many SEO’s have been looking at additional metrics over the last few years to supplement (or in some cases replace) rank tracking.

Here are the 10 metrics that I try to analyze when the data is available:

1. Organic Traffic Volume (Non-Branded in particular)
2.  Keywords Driving Organic Traffic
3. % of Organic Traffic from Non-Branded Keywords (vs. Branded)
4. Conversions/Revenue from Organic Traffic (Non-Branded and Branded)
5. Highest Converting Organic Keywords
6. Rankings for Top Keywords (Compared to Conversion Metrics)
7. Google PageRank, or Domain Authority (a Moz metric, similar to PageRank)
8. Quantity, Quality, Relevancy and Diversity of External Links
9. Quantity, Quality, Relevancy and Diversity of Linking Root Domains
10. Cost Per Conversion from Organic Visit

There are others, but these are the ones I’ve been using for clients where I have correct tracking and sufficient data to use them.

Seems like a great list, right? Sadly, it WAS a great list, but as time has passed and as Google has made changes (cough, Not Provided, no more than 10% my a**, cough) and as search behavior has shifted, every point of data on this list has become massively flawed. Here’s the list again, but with explanations for why each metric now lacks value:

1. Organic Traffic Volume (Non-Branded in particular) – Tracking issues, lack of referer information, and Google’s Not Provided segment in particular have made this impossible. Not Provided prevents you from accurately determining what traffic was branded and what was non-branded, and while some tools try to overcome this limitation by assuming that the keyword data that is still available is an even sampling of the whole, they fail from the perspective of statistical accuracy.
2.  Keywords Driving Organic Traffic – Once again, Not Provided makes this impossible to gather accurately and Google Webmaster Tools data is believed to be highly inaccurate as well. There is no way to determine exactly which keywords drove organic traffic and in which amounts with accuracy.
3. % of Organic Traffic from Non-Branded Keywords (vs. Branded) – Same issue as #2.
4.  Conversions/Revenue from Organic Traffic (Non-Branded and Branded) – This is also obfuscated by Not Provided. Beyond that, almost all conversion funnels should be taking multi-channel attribution into account and without that, you’re left with last click conversion data, which isn’t an accurate representation of your actual conversion funnel. If you need some convincing on the value of multi-channel attribution, go read some of the awesomeness that is Avinash.
5. Highest Converting Organic Keywords – Same reasons as #4
6. Rankings for Top Keywords (Compared to Conversion Metrics) – Sadly, rankings are still one of the few metrics we can track with any accuracy, but it’s an incredibly flawed one because of how different each search result page can be for each user (based on location, search history, signed in or signed out of Google, intent, SERP features, and etc.). While we can still track rankings, the massively different SERP’s make it impossible to apply any sort of CTR estimate to a ranking (well, you can to 15% of SERP’s, but yeah, that’s a small slice), and Not Provided borks any attempt at conversion attribution.
7. PageRank, or Domain Authority (a Moz metric, similar to PageRank) – PageRank, as made available to users, is so rarely updated that it has little, if any, value as a metric. Sadly, I don’t put any trust in any of Moz’s metrics for two simple reasons: the size and scope of their index changes constantly and they have a tiny index compared to Google, a small slice of the overall pie. While link building efforts should, in theory, increase your site’s Domain Authority, it could just as easily rise and fall based on the depth and breadth of Moz’s latest crawl.
8. Quantity, Quality, Relevancy and Diversity of External Links – While we can attempt to track this via numerous tools (Moz, AHrefs, Majestic SEO, and etc.), none of these link indices have all of what Google has. They don’t crawl everything and they don’t re-crawl everything, at the same rate, depth or breadth that Google does. While these tools can give you a decent sample of your overall data, the accuracy and value are questionable.
9. Quantity, Quality, Relevancy and Diversity of Linking Root Domains – Same reason as #8.
10. Cost Per Conversion from Organic Visits – Same reasons as #4.

OK…So Now What?

For the right company, in the right space, with all of the right tracking solutions configured in exactly the right way, getting most of this data might be possible (I say most, because you’re simply never going to get it all). I have never, in all my years in online marketing, encountered a company that had every path into and out of their conversion funnel tracked correctly.

Some people will have their web analytics implemented perfectly, but then they’ll have a phone number on their website that generates offline conversions and has no call tracking. They’ll have web analytics and call tracking implemented correctly, but their sales cycle takes 6 months and they don’t effectively connect that conversion 6 months down the road back to the online data they gathered previously via a good CRM.

There’s a great article here that goes into some of these issues in more depth, if you’re curious. If you’re up for the challenge, you can get to some of this data with custom reports, custom dashboards and advanced segments in Google Analytics, assuming all of your tracking code, event tracking, and conversion tracking is in order.

Overall though, there’s just a clear lack of good clean data from both sides, and that’s an enormous issue when it comes to picking and tracking metrics that actually have value.

Bad data leads to bad decisions, so if you know that the data you’re using for a certain metric isn’t reliable (and let’s face it, it probably isn’t), then you really shouldn’t be using that data to make business decisions, much less as a measure of SEO success.

So, if none of the above metrics can be trusted any longer, and I strongly believe that they can’t, then what are we left with? A problem, and one to which I haven’t yet found a good solution.

If you have a solution, or another SEO-centric metric that you’ve taken to tracking and that you’re confident in, or the world’s most awesome SEO-centric advanced segment, please share it in the comments!

Sam McRoberts is the Director of SEO at Point It by day, and the owner of VUDU Marketing by night, and has been involved in the online marketing world since 1999. He’s a foodie, a gamer and a traveler, and loves all things entrepreneurial. You can connect with him on Twitter, LinkedIn and Google+

Latest posts by Sam McRoberts (see all)