Forum Moderators: open
"G is determining SERP via click through rates based on their toolbar results along with what they feel are quality pages related to the subject."
I am not really sure about that, in my case I have sites having very high traffic but my site which was #5 for a long time for a highly competetive keyword has dropped to #19 after the backlink update, and the sites which are before mine are not that great in terms of traffic or brand name...
I am still trying to figure out what are the changes that have affected my listings, though what I could say for sure is G is not considering traffic or CTR at this time, if that was the case my site should be in top 3.
Cheers!
Wasn't there a long thread in the supporters forum a while back (I have no time to look for it) in which Brett speculated that they could do this and GG explicitly said they weren't?
I tend to agree with chris based on what I have heard (second hand) from employees at google, although wether it's 'CTR' or 'SESSION' behavior (users returning for not finding what they wanted) - they are closely related.
The thought that google would use CTR to rank a site seems a bit off to me. They might DO that in essence by allowing these metrics to determine the viability of an algo, but to do so on an individual keyword would not give them a good sample size at the rate that they change algo's for the vast majority of keywords.
how do you guys take those quotes and table them?
Click on "style codes" link that is on the left the next time you post. It will tell you about [ quote ] [ / quote ]
how do you guys take those quotes and table them?
[qu*te]how do you guys take those quotes and table them?[/qu*te]
From: Reply to this topic > Style Codes
Certain categories (webmasters) of people use the toolbar more than others (neophytes). This would skew the results too much.
They would have to rank the sites according to specific searches, and people search on all sorts of odd phrases.
What they click on is not necessarily a good indicator of how good the site is, only how good the snippet is.
How long they stay at a site is not a good indication of how good the site is. Sometimes the search is so good that I get the information I need from first glance.
Just because I go on to look at more sites does not mean that the current site is bad. Maybe I just want to make sure that the current one has the best price.
It would just be a bad idea to base searches for me on the CTR of other people.
I believe the fluctuations are the result of differently tweaked algorithms. Google randomly uses one algorithm or the other each time a search is made, the data is passed back to the Google Tech's and they make judgements based on how each algorithm is performing. This way they can try out several 'tweaks' a month rather than just one. It also stops webmasters from being able to research algorithm patterns and therefore spam the index.
How do you know this as a fact, and if they find it important enough to take the time to consider, then why wouldn't it play a part in SERP?
This is actually a very complicated question.
First the how do I know it as a fact
1) I don't know it as a fact
2) It is something I am fairly interested in
3) I think I know how google thinks. I could be wrong or arrogant, but I feel like I have a good handle on WWGD.
4) I have had conversations with people that I know work for Google (none of which have specifically confirmed what I am saying)
Why wouldn't it play a part in the SERPS
1) I have always felt (as you do) that the toolbar data would be very valuable to google. My sources indicate that google has been less than pleased with this as as something that can be put to good use.
2) Having a high CTR is a good indication of giving the user something they are happy with as far as a good SERP.
You raise a valid question as to why this wouldn't be the same thing as putting it in the scoring.
CTR in and of it self actually tells you little about the VALUE of the page, but tells you more about the QUALITY of the SERP when measured against a bunch of them.
This may not seem to make sense at first. Aren't thhey the same thing? No. Especially if you start FACTORING CTR into the results themselves. What would happen is that pages with good CTR would get higher placement that would give them higher CTR and the process would repeat itself (yes there are ways around this).
Also - Google tends to only putthings into their algo that influence the quality of the SERPs. I have yet to see anything that suggests that CTR really is a good metric for a quality page.
If it is - then it would have to be as good as something else that google would then leave out.
I think when they do the CTR - they basically are looking at changes and satisfaction. They don't want people leaving google.
So that if they change an algo and result x used to get x% CTR - and now it gets half that - they aren't happy. Of course you could argue just by putting high CTR stuff at the top - they would accomplish the same thing.
It certainly is a valid point - and I can't really argue that heavily against it - other than I think:
1) It wouldn't improve the QUALITY of the SERPs
2) Because of this it would be un Google Like
3) They don't really do it enough to help with INDIVIDUAL results - but it would help with measuring changes in the SERPs.
Here is some more info from GoogleGuy:
"We normally don't track redirects on urls because it slows users down. That data is useful though, so we sometimes do random sampling to make sure that our quality is still high. If I had to guess, the sampling has already finished for you, Brett."
and
"We do our best not to slow users down. But we also do our best to measure our quality. So tracking redirects a small amount of the time seems like a pretty good compromise. Like oLeon says, this is something that we've done for several months or more. Definitely folks should drop an email to google if you think there's a better way to work it. We like getting good suggestions, and webmasters give really good suggestions most of the time. :)"
[webmasterworld.com...]
"rfgdxm1, every so often we collect clickthrough data to see how we're doing on quality. We don't do it most of the time because it slows users down, and we hate to slow our users down. So it's a simple explanation; no need to invoke the CIA or anything. ;)
Also, do you remember the conversation we had about picking on Google when other sites need it more? Most other sites use redirects a lot more than we do. ;)"
[webmasterworld.com...]
a bunch more here:
[webmasterworld.com...]
Here is the official line by the way:
"When Google uses a URL redirecter, if you click on a URL from a search result, information about the click is sent to Google, and Google in turn sends you to the site you clicked on. Google uses this URL information to understand and improve the quality of Google's search technology. For instance, Google uses this information to determine how often users are satisfied with the first result of a query and how often they proceed to later results."
None of this specifically precludes what you are suggesting. I don't think if google DOES do what you are suggesting - you are going to get anyone from google to confirm it.
Oh - and yes I realize that the toolbar is slightly different scenerio than the above. Notice that google puts no codes at all in the SERPs most of the time - which would make measuring CTR from SERPS EXTREMELY difficult - if not impossible the way that googles results fluctuate so much.
<SCRIPT>
<!--
function ss(w){window.status=w;return true;}
function cs(){window.status='';}
function clk(n,el) {if(document.images){(new Image()).src="/url?sa=T&start="+n+"&url="+escape(el.href);}return true;}
//-->
</SCRIPT>
It certainly looks like the way they used to encode urls. Well that shoots some of what I said above, but again - they must have just started doing this.
It certainly looks like the way they used to encode urls. Well that shoots some of what I said above, but again - they must have just started doing this.
They've been flitting with the idea for some time and it's randomly shown up on the search results. It became a permanent feature on the 30th March 2004.
They made a mistake on the day of release and the tech's accidentally put the redirect code straight into each URL which meant that you would see it when you hovered over the link. They changed it within 4-5 hours and no-one seemed to notice.
I guess they don't want to much speculation on how they are going to use the data so are keeping it hidden. Sorry GoogleBoys if you didn't really want it pointed out but I thought it was relevant to the thread.
Published 9/5/02
"Methods and apparatus for employing usage statistics in document retrieval"
"Methods and apparatus consistent with the invention provide improved organization of documents responsive to a search query. In one embodiment, a search query is received and a list of responsive documents is identified. The responsive documents are organized based in whole or in part on usage statistics."
[webmasterworld.com...]
A little off topic I recall when Yahoo took click and return time into account as a result websites started putting in redirects so that people couldn't use there back button, one reason that I believe that rankings wouldn't be determined entirely by clicks as it would be too easy to manipulate.
one reason that I believe that rankings wouldn't be determined entirely by clicks as it would be too easy to manipulate.
Agreed, with the amount of spam clicker software that could be modified to boost rankings this would be a very delicate move by Google.
I suspect that it is still a random feature, but being used more now.
I'm on the west coast in the USA and I don't get it. Just some good old regular <a href=>.
Immediately after each link you'll see a javascript call code onmousedown="return clk(7,this) pointing to a javascript section in the head tags. Do you see this?
If those are the 3 pages that the algorithm favours its only natural they will be your most requested pages in general.
The fact they get more impressions is relational to your search engine positions and nothing to do with the pages popularity.
In my opinion, you can't judge the quality of a site until you have clicked on a result anyway, so to penalise sites with low a lower click through rate would be wrong.
Are you sure that you are not looking at the code for the ads?
I should have worded that differently. The 3 pages I was referring to are pretty static in SERP's. I meant other pages on my site that are constantly being moved around, and I mean dropping way down and then returning back on top again. At first I thought it was just G refining it's update. But as those 3 pages have remained unchanged and the others keep jumping around, I figured it had to be something else.
Jpell
Like I said, the only ones that uses javascript are sponsored links. Onclick, onmouseover and onmouseout.
Are you sure that you are not looking at the code for the ads?
Okay, here's an example, I hope it is in keeping with the TOS of WebmasterWorld:
On an actual search for 'widgets' the second natural search result is for widgets.com which if you look at the coding has this link:
"<a href=http://www.widgets.com/ onmousedown="return clk(2,this)">www.widgets.com/</a>"
and then in the head tags it has this code:
<script>
<!--
function ss(w){window.status=w;return true;}
function cs(){window.status='';}
function clk(n,el) {if(document.images){(new Image()).src="/url?sa=T&start="+n+"&url="+escape(el.href);}return true;}
//-->
</script>
which give the link to widgets.com a Google redirect link that is hidden from the user.
Can you see that or not?
I do a search, so a view source and search on "onmouse" and the only places that it finds are the ads.
In the past I have had tracking turned on during one of my sessions, and I have also found searches where it seems like it is always turned on. But it quite simply is not the case that they are tracking all SERPs for everyone right now.
I suppose that they just might be spying on you personally, or maybe it is only highly competitive phrases. It could be your OS or browser of choice. Maybe they are interested in what people from your ISP or your town are searching on. But they are not tracking me or the searches that I tend to do.
I checked just now and it is in the main serps too on my computer. While my positions on most of the keyphrases are quite stable, for the most competitive keyphrase I am in, my page is ranked between #5 to #24, moving after a few hours or so. Most of the other sites in the top 20 are in similar situation. This has been going on for the last few days. My suspicion is that click throughs are being used.
