Recalculating the Click-Through Rate in Search

Much has been made during the past two months about Google losing its fastball in search. Numerous news articles have detailed a lack of consistency, a dependency on content “farms,” and a general dissatisfaction with the experience coming from search with the undisputed category leader. In response, Google has unofficially announced war on those taking shots at the algorithm by implementing changes in an effort to make search more personal, less content dependent, and even more social.

Regardless of the quality changes being made by Google, Hitwise recently released a statistic that caught my eye. It is not a new statistic, as Hitwise has reported on it for a long time. But its implications have not been fully discussed. The statistic is success rate. Simply explained, it’s the percentage of users that actually left the site after doing a search. We constantly talk about the power of search being the ability to meet consumer intent with relevant content. In theory, this number is one way to do so. It would be a more worthy metric if it measured actual clicks on a link on the result page, but it can be a directional jumping off point, at the very least.


There is an inherent problem with this in that more and more searches across all engines are designed to provide information without requiring a click or departure from the site. So, there must be a factor of variance allowed due to people who might type in a keyword such as “weather” and find the local forecast without having to ever leave Google. But, even with those informational queries being met with immediate response, the variances between Bing and Google are surprising. Even more interesting is that the number from January 2011 represents a sizable improvement for Google where it was hitting on less than 60 percent of all queries leading to a click as recently as October.

This data raises at least the semblance of an interesting debate for advertisers on how to think about the success of their paid search program. To date, brands think of the metric “click-through rate” (CTR) as a zero-sum game. Either I got a click or I didn’t. If I did, then we put it in the successful fire category; otherwise it goes in the dud column. However, this metric has been built on the premise that if I didn’t get the click, then someone else on the page did, and therefore my click-through rate suggests that if I had a 5 percent hit rate then 95 percent of all other clicks went elsewhere.

This data says that’s a flawed thought. If one-third of all queries led to no clicks, and, being generous, half of those were informational in nature, then we still have a 15 percent no click due to a failure to properly connect intent and content. If this is correct, then brands would need to do two things:

  1. Create a new metric – eCTR – which measures their actual CTR when someone clicked somewhere on the page. This serves two purposes in that it allows brands to know when they really were selected over other choices and also to know when everyone failed to connect.
  2. The problem with option one is that to do it right, there would have to be some way to properly track impressions served and receive that data from the engines. The engines themselves would have to provide the amount of impressions that failed to create a click for everyone. However, simply allowing third-party ad serving on search impressions would at least allow brands to modify their messaging to better target when they did fail, whether a click went elsewhere or not.

In a recent blog post, Google’s Matt Cutts takes issue with the positioning of this data and what is a “successful” search. In my opinion, that is changing the conversation away from where real value exists. Brands have a very simple definition of a successful search – one where a user clicks on their brand site. If more than 30 percent of all users are just leaving the search results page without a click, then that moment is not successful for brands, yet they have no way of improving that experience for users based on the material data.

There is not a major search player today in the United States market that has shown any substantial interest in addressing this problem. So, while search engines talk about optimum consumer experiences and creating meaningful connections with brands, it seems they would rather debate the definition of a third-party metric than give brands the insights and ability to impact their own true click-through rate. Until that discussion comes to the fore, brands will be left looking at a number that they have to question. And by anyone’s definition, that does not equal success.

Related reading

Overhead view of a row of four business people interviewing a young male applicant.