Google's Ranking Metrics That Make a Difference for SEO

Following on from Matt Peters' great post over on SEOMoz, I thought it might be worth recapping on the background of a key metric that currently shifts the needle for our clients.

2011 has been a really exciting year for Search Engine Optimisation (SEO), with much more valuable ranking methods coming to the fore. In particular one of the much underhyped metrics has come to be (almost) top dog in the list of metrics that drive SEO performance: SERP Clickthrough Rate.

SERP Clickthrough Rates & SEO

Well understood in Pay Per Click (PPC), clicks from the search page have been discussed by the community long before Google's patent on the method was published.

The idea that Google would monitor CTRs to identify which organic listings get most often clicked on - and which that are clicked on result in fewest return visits to the search result page - to highlight better quality pages for the term searched for is, it seems, obvious.

For a long time however, discussion erroneously centred around Google's statement that Time on Site (ToS) or other metrics they would be aware of from Google Analytics are not used in their ranking algorithm. Of course, while it is true that Google Analytics data is not used, it's also true that such data is not required to understand if a searcher session returns to a SERP after clicking on a result: the data is there to be observed directly on Google's own pages.

It's also even easier to see how that behaviour could be aggregated across multiple synonyms for the same URL, to provide a better correlation of signal and so provide a solid filtering component to the main algorithm to drop out poor quality pages.

A big part of the stated aims of Google's Panda / Farmer update can be achieved by adjusting the weighting of this part of the algorithm for certain sets of terms - particularly non-transactional terms and the 'How To' content farms - that in fact provide no such service - which suffered as a result of the update.

A big element of scepticism is required when drawing conclusions from the patents Google publishes and speculating on the core components of Google's algorithm, but at QueryClick we've consistently seen work on improving the matching of landing page to relevant search terms with improved snippet relevance pay off in a big way for long term ranking performance.

Using Google, Yahoo! & Bing's adherence to the NOODP and NOYDIR robots metadata, and by including a landing page's target keyphrase term in the meta description & title tag allows us to completely control the snippet generated for a page in a SERP.

Working with these components to design our organic listing with the same care lavished upon a high cost per click PPC listing, we can greatly improve clickthroughs for the term even while holding the same rankings. We also see over time an improvement in the listing performance as a result of this work though again, we are making improvements to keyphrase matching and freshening content which also improves performance.

Where the correlation becomes stronger is in the ongoing performance improvements that occur after the changes have been made and the listing generates a history of 'Good Searcher Behaviour' (GSB, anybody?).

We can then extend our work from simple ranking improvements to conversion benefits by matching the landing page content better to the snippet's sales message (or other call to action), thereby improving traffic conversion for a traditionally high new or first-time visitor traffic source: greatly improving new growth sales for the client.