For web marketing training, presentations and workshops visit Pacifica Training. If you wish to ask questions about website marketing please head over and register free at the website marketing forum, a knowledge base and community driven discussion group. There's also an opportunity to earn extra money working from home which is part of the tourism database we've created called Touristino. For local news visit our new Community Hastings website.
The bounce effect
Published: November 17, 2006
Over the course of 2006 I have been working with SEO and I'm seeing a pattern emerge in Google. It's a pattern in the roller coaster sense and I'm wondering if it's a symptom of too much SEO.
First a web site or page ranks well. This is known as the honeymoon period. I've seen this first hand as it has happened to many of my articles. Once the high rank period has expired, Google scores the page, judging its worth by studying the bounce effect. It's this bounce effect that determines the initial stable ranking position of the site or article. To increase the potential rank of the site, we need to know what the bounce effect is and how we can get some on our sites.
What is the bounce effect?
The bounce effect is not something we want, it's something we don't want. To sum it up, bounce is a measure of the time it takes a visitor to click on your page in the Google results, read the content, then bounce back to Google results. Google knows you, you see. It knows your IP address and session variables. These are means of tracking you as a user. If Google sees you find a page and return within seconds, it will assume that the resource served isn't what you're looking for. In return, this lowers the rank for the site and that given phrase. If the bounce effect is substantial enough, you'll plummet in the results. If Google doesn't see you in a while, it will assume you've either spent a lot of time reading good content or have been signposted to resourceful areas. This will increase your rank.
The bouncing pattern
The pattern I'm referring to is the cyclic motion of the result sets. I'm seeing pages, even sites, go up and down several times a month. Currently stability is poor and questions are being asked as to why sites do well, drop, then do well again.
The theory
There's too much SEO. Manipulation of the results is something that the search engines are driving millions into averting. Their aim is to maintain and serve quality results. They say you cannot pay to manipulate the organic listings but you can if you employ a good SEO consultant. I'm sure Google are working hard to counter this. Since there is plenty of content rich information that is optimised, I'm guessing Google is cycling the results and using the bounce effect to come to a conclusion as to the final rank of the resource.
Social networking is the best measure of quality. It's consumers who best decide if the information served is right for them and social networks such as Digg and Del.icio.us offer a great way of determining this. It would make sense for the major search engines to emulate this voting system to judge quality. What better way than employing a bounce effect?
Final thoughts
What all this means is nothing new. We need to focus our content solely for the user but we already know that. Unfortunately, search engines aren't that wise yet so SEO will still have a huge part to play, and to a degree, will always be with us, unless of course Google becomes a social network too.
I can't see an end to this either, in fact, I can only see growth of the SEO industry as different but fortunately completely transparent methods are being used to judge rank. The hard part is staying on top of this and compiling a user base of statistics to back up the research, which I'm sure is time enough for Google to shift the posts yet again.
Useful?
Assistance?
Comments
Online TV John says:
WebGeek says:
Great post, Ed! I think you may be onto something here. I've observed this many times. It's not difficult to install tracking like that even if people aren't logged in. Like you said all you need is the IP address and session id. And, they could pull data from Google Analytics as well. That would be a sly reason to offer a tool like that for free. Hmmm.
SEO says:
I tested this in 2003 when it first got bought up. We would conduct a search, click one of our test sites on page 1 then hit the back button in under 10 seconds and visit test site 2 which was also page 1 and browse for 10 minutes or so simulating it was an informative resource.
Unfortunately we didn't get a change in SERP's for either site.
SEOchat also just done a test on this, and also on the bounce rates indicated in Google Analytics using about 50 different people in the test.
They did detect the test sites with high bounce rates fall in SERP's, and the sites with low bounce rose up in SERP's.
The trouble with this is the 100's of variable used to rank a website, and the constant state of flux in the index it's so difficult to test with 100% conclusing results.
Google are sneaky buggers, they do have in one of their many patents about "using visitors browsing and bookmarking habits to determine the rank of a webpage"
This clearly says to me they are either working on it, or it's allready a small factor in their current algorythm. No doubt their ranking methods will shift from factors that can be manipulated (ie backlinks, paid links, on page factors etc) to other ones that are alot hard to manipulate like user browsing habits and different trends for a webpage.
Good post Ed
Do you have stats to prove this, it all seems a little complicated and I'm guessing you would have to be logged in to your google account to have these logs recorded.