quiet_man - 12:41 pm on Dec 21, 2009 (gmt 0)

When you add up all the points mentioned in the replies above I wonder whether the idea of a set formula or table correlating rank with CTR has had its day? The landscape in 2006, when the AOL data was leaked, is very different from what we see today. In any given SERP, any one of these points can throw the formula completely:

* Universal Search [webmasterworld.com]

* Blended Search [webmasterworld.com]

* Real Time Search [webmasterworld.com]

To me, its not just the presence of any of these elements in the SERPs, its their prominence. Your site can sit at #1, but if there's a big map or images or a scrolling ticker at #4 then the searcher's eye is immediately drawn down the page. In that case, #1 spot might not even get the highest number of clicks.

In addition to these elements, as mentioned above each of the following can have a major affect on CTR and so may trump actual SERP positions:

* Site Title

* Meta [webmasterworld.com] Description [webmasterworld.com]

* Sitelinks [webmasterworld.com]

And maybe the final nail in the coffin:

* Personalised Search [webmasterworld.com]

Privacy issues aside, the point about personalised search (custom-search-by-default) is that we just don't know what position our site occupies in the SERPs that any one searcher is looking at. So while we search marketers can look at a set of un-personalised SERPs and see our site at #1 (or #2, or #3, etc), we just don't know how many other searchers are seeing this. How can you create a table or formula if you can't be certain about each SERP position?