« iPhone Compassion International | Home | NeO Blog Moved! 301 »
Is CTR , PageViews and Average Time On Site affecting SERP’s?
By rumblepup | June 11, 2008
Brian Mark is currently in Chicago attending a conference, so David and Cshel where kind enough to ask me to fill in on SEO 101 . We where inspired to talk about the recent speculations on how CTR (click through rate) Page views and Time On Site can be a factor that Google is using to determine search positions. This topic came up because we caught an article on sphinn from Bradley Spencer .
Bradley discusses the possibility of Google using CTR as a factor of their organic listings. He came up with some very interesting points that do make a lot of sense. Google is in the business of quality search results, and to do that, Google has created a plethora of different ranking models to keep enhancing their organic results. Being that we are talking about some of the most powerful search technology in the world, it’s not far fetched to consider that not only can they track click through rates, but bounce rates and average time on site stats as well, and have built into their algorithms a ranking model that uses this information. On the surface, this User Experience ranking model should be a great metric. The simple reasoning is that by monitoring what we click on in the search results, Google can determine if a site is drawing the amount of clicks it thinks that position should hold. The more clicks a site gets, and the more time someone spends on a site, or page views, the more it’s showing that it’s performing at a higher position, so they would increase it’s placement in the SERP’s.
Danny Sullivan discussed this at length , adding as well that personalized search is a big result of this algorithmic shift. Shoemoney also covered this trend when he considered the future of SEO .
In our discussion, I brought up some of the problems I have with this kind metric. Click through rate and bounce rate are two factors that are pretty subjective, depending on the type of traffic you’re trying to get. In my opinion, using these types of factors can be a slippery slope, because user intent is always a hard metric to gauge. Let’s say that Google uses bounce rates as a determinate, and sets the time to bounce at 30 seconds or less for a “bad” site, and varying degrees of time on site for different “grades.” 30 seconds is a lot of time, much longer than you really think it is. In fact, you probably read this last paragraph in less time than that. There are excellent sites out there that have created content that is spot on, you can get what you need, and be off in less than thirty seconds.
On the flip side, a slower bounce rate from a site can mean 1 of 2 things. The user is so enthralled with your site that they want to see more and more and more, like my wife and Etsy , or, your user is NOT finding what they want, and keep searching the site. That latter category is a killer for e-commerce sites because sales cycles can be long, and the information you glean from these “slow” users can lead to more questions than answers. So a site might be getting a SERP benefit from clicks and pageviews, but the quality of this traffic is probably not a factor. User experience is as unique as humans are. A two minute visit to a site might be fantastic for one person, while it’s a massive disappointment for another.
Furthermore, I think it’s a bit easier to game a system with CTR as a type of metric. Let’s go back to the days of overt click fraud (not that click fraud is gone, but as far as Google is concerned, they’ve gotten much better and weeding out bad clicks) Back then, it was easy to buy a bunch of clicks form “click teams” or “click markets.” So let’s apply that to this metric. If it actually gains sufficient value, there will be a whole new underbelly network of doods clicking on SERP’s to increase the CTR. Just as they are trying to fix the page rank and link methodology, if adding this metric, they are adding a whole new set of problems to their plate.
Carolyn (cshel) brought up a good point as well about bounce rates. In the case of a MFA site, what’s the performance we’re looking at here? If you click on SERP that leads to page with prominent advertising that you are actually compelled to click on, is that indication of bad traffic or highly converting ad placement? A high bounce rate might mean high performance ad clicks. Google is not a charity organization, they are happy to make money as well. The high ad CTR might determine to them the ad placement on the page is successful.
And speaking of GoogleAds, does the traffic generated from ads figure into this CTR as well.? According to my experience, as well as NeO’s and Cshel’s, the Google Ad Team and the Google Search Team are kept separate from each other, sequestered even, like early monks copying bibles and literature by hand. (Insert funny picture of Matt Cutts in monks’ robes and a shaved pate)
On the web, there’s two ways to get traffic, you earn it, or you buy it. Heck, Google sells traffic.
Although I love the idea of User Performance as a metric, it’s important to gauge what valuable user performance is. Simple CTR and Average Time on Site might not be enough evidence.
David wants to try and experiment with the CTR scenario. He’ll be posting up his results when he’s done with the experiment.
You can listen to the full podcast at ClickThrough in Organic
Topics: SEO |
June 11th, 2008 at 2:50 pm
Hey there! Thanks for mentioning my article on whether or not Google uses CTR to figure into their organics.
I really like the layout of your site…. do you Sphinn also?
Best, Brad
June 11th, 2008 at 3:02 pm
@Brad the Wordpress SEO: We all have sphinn accounts (Brian Mark, Cshel, NeO, Rumblepup) the whole gang… Great article and thanks for the friendly graphs!
We also discussed the post for almost an entire show! I’ll come back and post the link to the podcast as soon as Brian gets back from the IRconference.
June 11th, 2008 at 5:37 pm
WOW! I’d LOVE to hear that podcast…
Thanks, Brad
June 18th, 2008 at 10:22 am
I have tried to start discussions about this on many forums, usually ignored or shot down.
On its own CTR and bounce rate are not great for the reasons make in your post, but when you combine it with other factors it becomes much more powerful. For example, if a page has a low ctr and a high bounce rate for a particular search term or intent then down it goes.
Cheers,
Tony
June 18th, 2008 at 12:18 pm
@Tony: you might then enjoy the podcast where we covered this topic - you can listen to it in streaming or download it here http://podcast.neo1seo.com/2008/06/11/google-clickthrough-in-organic/
db
June 18th, 2008 at 1:47 pm
We’re probably gonna look a lot more into this. As I said in the post, David is gonna be running some tests, so we can see if there is a definitive result for this.
June 19th, 2008 at 4:02 am
[…] not a lot of information is available about this. It is being discussed by SEO experts, but it is still seen as a future trend, currently in experimentation […]
June 19th, 2008 at 9:23 am
[…] across Neo’s blog where he is discussing the subject and promising to do some experiments. See: Is CTR , PageViews and Average Time On Site affecting SERP?s? | Top SEO Consultants __________________ Search engine marketing for beginners and a "do-follow" incentive […]
June 19th, 2008 at 9:43 am
Hey David. You could do with the subscribe to this post widget so that commenters get e-mail updates when a new comment is posted.
Cheers,
Tony