Your sites have been audited, improvements have been made and there is more to come in the road map. The technical, editorial and marketing teams are trained in SEO best practices and are integrating them into the daily workflow. Sounds good, but are your efforts working?
That’s where tracking and measurement comes in. When it comes to SEO publishers need to understand and effectively communicate progress and success.
SEO is a never-ending process and without good benchmarks, clear and realistic goals and solid reporting it is easy to lose momentum within the organization.
So what are the best ways to track and measure success?
I’ll leave the in-depth answer to the analytics experts and Excel ninjas out there (along those lines a good recent post is Ian Lurie’s SEO Analytics, Middle Earth-Style).
In this post I’d like to highlight some of the key metrics for publishers to focus on.
Primary Success Metrics
- YoY natural search referrals – this simple and straightforward metric is the #1 measuring stick for news and content sites. It is best to look at YoY growth since month-to-month can be too easily skewed by seasonality. YoY and rolling three months are a good combination.
Remember though that YoY can also be strongly impacted by the news cycle. When things like Tahrir Square or the Royal Wedding hit, if the same month one year later does not have the same traffic levels that isn’t necessarily a negative sign. When it comes to analytics everything is relative.
- Branded vs. non-branded – publishers with well-established brands tend to pull in a lot of branded searches, so separating non-branded referrals is a better way to judge your overall content optimization efforts. However you want to measure branded referrals too since this is a way to monitor growth in brand awareness and popularity (which can also have an impact on SEO).
- Search traffic by site section and content type – to make the data more meaningful it is important to segment it both by channel/section and content type (articles and blog posts, features, slideshows, video, etc.)
- Total number of pages receiving search traffic – looking at the total number of pages that receive at least one visit from search is a decent way to measure total indexation, how older and archived content is performing and how well the site is doing for longer-tail queries. SEOmoz covered this a while back in a post on indexation for SEO and I thought it was a good idea.
- Total number of referring keywords – this is less insightful than it used to be due to the Google secure search “keyword not provided” issue, but it is still worth capturing this data.
- Top referring keywords – this is also less insightful now but still useful to examine, both for the site overall and for specific sections and high-priority content. The search queries data in Google Webmaster Tools can be used to offset “keyword not provided” deficiencies in Omniture or Google Analytics.
- Pageviews, time on site, bounce rate, etc. – searcher intent and behavior varies depending on the type of query and content (as well as the individual). In some cases search traffic is more likely to be “one and done” since a successful search outcome means the user found what they were looking for; in others a multi-page visit might be more the norm.
Regardless of the circumstances, you want to do what you can to keep visitors on the site. These metrics help you to measure that and to compare search with other types of site visitors (like social media and direct navigation).
- Conversions – this comes into play much more in fields like e-commerce SEO, but publishers have desired conversions too. At a basic level this might be print, email newsletter or RSS subscriptions. Getting search visitors to click through to a second piece of content could be another. It can even be things like increasing follows/likes/+1s to help with social media efforts.
Secondary Metrics
There are a wide range of additional metrics that can be evaluated, in particular things that do not directly define search traffic or searcher behavior but do have an impact on SEO success.
On the technical site this includes tracking Google and Bing Webmaster Tools errors combined with the crawl data from enterprise SEO toolsets. This provides a way to measure the progress of technical fixes and improvements and a way to identify new issues that arise (as they always do).
For editorial teams this can be a roll-up of feedback based on periodic spot checking: how effectively search-friendly language is being utilized, how well key page elements are being optimized, the depth and volume of content around important topics, etc. Such things are more subjective and harder to report on comprehensively but there are still meaningful insights that can be gained.
In the marketing category secondary metrics include tracking domain authority and page authority (for key pages) over time, along with internal and inbound links, social shares, additional brand signals, etc.
What about Rankings?
It is hard to talk about SEO analytics without any mention of rankings, but particularly for news and content sites it should not be a primary success metric.
The increase in personalized search rankings (for both signed-in and not signed-in users) combined with QDF (query deserves freshness) and the Google Freshness update, which give a temporary boost to brand new content around hot and trending queries, means that everyone’s rankings are a bit different, and rankings are frequently in flux. So it is not possible to provide truly accurate ranking data.
For these reasons it is better to move away from rankings and focus on YoY and rolling three-month natural search referrals (branded and non-branded) as the primary success metrics.
This can be supplemented with the available keyword data from Omniture or Google Analytics plus Google Webmaster Tools and third-party tools like SEMRush to gain general insights into what types of terms are bringing in search traffic.
If you still want to monitor rankings for a select group of high-priority terms that is fine, just remember the limitations and make sure everyone else understands them too.
Reporting Format
One last tip: how you share the data within the organization is nearly as important as what you are tracking and measuring.
There is a lot to be said for clean, simple, visual reporting. The main thing being that people may actually read it!
The best approach is to utilize a multi-tiered format since not everyone needs or wants to see all of the reporting.
Start with a simple one-page summary that conveys the top-line data in a clear and meaningful way. Then use additional spreadsheets to provide in-depth analysis for those looking for more detail or who want to drill down on specific things.
Sharing the basic reports across the organization can also help to foster a little healthy competition between titles and/or site sections, which in turn helps you to gain more SEO buy-in and execution.
What Else?
These are the primary and secondary metrics that I like to utilize. What else are you tracking and how are you reporting the data?
Let me know in the comments.
Nick Stamoulis says
Website owners tend to focus too much on rankings. What’s really important is visitor growth and conversions. Rankings fluctuate for many reasons and focusing too much on rank can drive you crazy.
Adam Sherk says
I agree Nick. There many better things to monitor and evaluate than rankings.
Derek Slater says
Hi Adam – many good thoughts here.
I’d quibble with your assertion ‘search traffic is more likely to be “one and done” since a successful search outcome means the user found what they were looking for.’ On my site, search traffic actually has better ppv and time-on-site stats than the overall site averages. There are a lot of factors behind that, but in B2B settings at least, the object of the search is often not a simple discrete piece of information, but rather information pertaining to a complex task or process. Search traffic is great traffic for publishers who know how to make the most of it.
Also, engagement can be viewed at a site-wide level as you note, but Google Analytics makes it easy to view engagement on an article-by-article basis. Publishers can improve the engagement/conversion of long-tail pages (pardon my jargon) over time as they see how searchers respond to those individual pages.
Adam Sherk says
Thanks for your feedback Derek; you make a good point.
I agree, intent varies from search to search (and person to person) so from the searcher’s perspective just what defines a successful outcome will also vary. And there can be a lot of variance between queries for breaking news and hot/trending topics vs. more evergreen queries, as well as between B2C and B2B content, etc.
And regardless of whether or not the person’s objective was met on the page they landed on from search, publishers want to entice the visitor to stick around and check out additional content (be that through relateds, touts for new/ or popular content, etc.). All of which makes metrics like pageviews per visit, average time on site, bounce rate, useful to look at.
I’ll clarify that a bit in the bullet point above.
Vahe says
Hi Adam,
Considering the current changes with Google Analytics and a shift towards analysing content groups/topics, what would you advise of some of the metrics/KPI’s that you are reporting on for existing client efforts.
Thanks. Merry Christmas and have a happy New Year!
Regards,
Vahe
Adam Sherk says
Vahe – sorry for the slow reply, I took a break during the holidays.
You’re right, top referring keywords and total number of referring keywords have lost most of their usefulness, though I do think it’s worth monitoring the Search Queries report in GWT for at least some partial data. Third party tools like SEMrush also provide a rough idea of the types of terms a site has search visibility for.
Re: content groups/topics, I think drilling by site section and/or content type has become more important now, as it helps to better focus on that at least to some extent.