View Full Version : does traffic influence positions?
inertia
12-07-2007, 07:58 AM
Do the search engines know and/or use the amount of traffic your site gets to influence positions?
In theory, yes. In practice, it is almost impossible to say. It also depends what source of traffic you are referring to. All major search engines know what the click through rate is for every result for a particular query. If the SE sees that for the query "blue widgets" item #2 is getting more clicks than item #1, the engine could determine that #2 must be more relevant and move #2 up.
If you are referring to traffic in general, only Google has the resources to track this in any sort of reliable way. There are three ways a search engine can know about your traffic. All major engines use one of these methods, Google uses all three. The first method is tracking hits to ads on pages. This is used by all major search engines, but does not give much useable information, because only a limited number of sites buy ads from that engine, so the SE can't get a valid comparison of traffic. The second method, used by Google, is to get a large segment of the web browsing population to install spyware on their computers that feeds data about every page they visit. There has been some research that indicates it is possible an increase in traffic from Google Toolbar users could cause a small increase in rankings. The third method, also only used by Google, is to offer a web site product to webmasters that sends all your visitor data and statistics directly to the search engine. This has the same limitation as ads in that only some sites will use the technology, however it is possible for the technology to become popular enough the data could be used to influence rankings. There is a theory, for example, that if Google Analytics determines that 90% of searchers who come to your site for a specific term immediately click a link to go to a particular second page, that second page may be moved up in the rankings.
Again, most of this is just theory, but it should give you an idea of what may be possible.
incrediblehelp
12-07-2007, 01:26 PM
Do the search engines know and/or use the amount of traffic your site gets to influence positions?
I am sure they do.
mjtaylor
12-07-2007, 01:37 PM
That makes three of us who agree. :D
And thanks to wige for explaining it so clearly!
A certain SEO blog did two unscientific tests about a year ago where they got two sets of pages indexed for relatively un-competitive terms, then recruited users to click the pages in different ways. One set of pages was used to test click-thrus on the SERPs - they had readers click only the page in the #3 spot on the SERP page for a specific query, and found in two weeks the page switched with #2. In a similar test, this time using Google Analytics, they found that after a few weeks of simulating a high bounce rate on a landing page the page actually started to slip in the rankings.
They did point out that the changes were slight, on non-competitive terms, and that there was no way to be certain the changes were not due to algorithm tweaks or other factors. I am hunting the link to the study and will post it if I can find it for anyone that is interested.
thyseo
12-07-2007, 09:07 PM
Before reading replies on this subject of traffic influece on rankings, I was 80% sure that the traffic helped keyword rankings. Now I am 93% sure that it helps rankings.
On SEO:
If you look at SEO after coming to wealthy peace of mind, it really is basic logic. The more we read SEO liturature, the more basic it gets and the more we understand what works for rankings.
Traffic is indeed in the popularity category and probably (93% sure) is included as a criteria for rank. Maybe Google uses Google Analytics to rank sites, who really knows?
Webnauts
12-08-2007, 10:57 AM
So search engines are becoming clones of Alexa? :lol:
jlmkanati
12-10-2007, 02:40 AM
Of course! traffics affects websites ranking from major search engines. traffics measures how site is established.
I don't think traffic influences rankings yeah clicks in organic ranking may improve the rankings but that also not proven.
So search engines are becoming clones of Alexa? :lol:
I imagine they would put at least some effort into minimizing the ability of user to manipulate the results. For top results in competitive categories, that would mean traffic variations on the magnitude of tens or hundreds of thousands of clicks to move a page one or two positions. But the concept goes back to making the search engines determine user intent - if nobody is clicking on what a search engine thinks is the best result, the engine needs to be adaptive enough to reevaluate what the searcher is trying to find for that query.
mjtaylor
12-11-2007, 07:01 PM
But the concept goes back to making the search engines determine user intent - if nobody is clicking on what a search engine thinks is the best result, the engine needs to be adaptive enough to reevaluate what the searcher is trying to find for that query.
Exactly.
And we know that Google's intent is to move toward personalization.
We also know that if you are a Google toolbar or gmail user that your click activity most definitely influences your organic search results. It is logical to extrapolate that Google uses the same technology outside of personalized search areas.
Cheers, MJ
NetProwler
12-11-2007, 11:26 PM
It is like saying people who consume large quantities of fiber in their food suffer lower incidences of cancer. The effect could have been <i>equally</i> due to their better overall healthy living trend. They may do more exercises, eat better and probably inherit better genes. You get the drift ?
The increase in rankings owing to a correlation in Google analytics is probably due to other reasons. Yes. Google and other major search engines could and do track the click through rates in the SERP and factor this into the rankings. But there is little evidence to suggest that they would be using Analytics data as part of this.
There are millions of sites who don;t use Google Analytics.
Webnauts
12-12-2007, 12:21 AM
It is like saying people who consume large quantities of fiber in their food suffer lower incidences of cancer. The effect could have been <i>equally</i> due to their better overall healthy living trend. They may do more exercises, eat better and probably inherit better genes. You get the drift ?
The increase in rankings owing to a correlation in Google analytics is probably due to other reasons. Yes. Google and other major search engines could and do track the click through rates in the SERP and factor this into the rankings. But there is little evidence to suggest that they would be using Analytics data as part of this.
There are millions of sites who don;t use Google Analytics.
AMEN.
And there are millions of people who do not have the Google bar installed into their browser.
Webnauts
12-12-2007, 12:25 AM
We also know that if you are a Google toolbar or gmail user that your click activity most definitely influences your organic search results.
Can you backup this valuable knowledge? I ask, because if this is true, then I can manipulate Google using similar techniques people use to increase their Alexa Rank. ;)
Can you backup this valuable knowledge? I ask, because if this is true, then I can manipulate Google using similar techniques people use to increase their Alexa Rank. ;)
As far as I know, right now this is in the theoretical stage. If you use the Google Toolbar while logged in, Google is logging every page you view, but they have not started using that information to change the order of results yet. This would not be subject to the manipulation you are suggesting, however, because the changes are applied on a per-user basis - for example, if I always click on result #3 for a specific query, after a few times I will see that result moved into the #1 position, and possibly highlighted (actually, it already is highlighted with the date or time I last clicked that result). This change in ranking however would only be visible to me - it would not affect the ranking of that site for any other user unless Google saw a major trend - a large percentage of users are preferring one site over another. In that case the rankings algorithm would try to adapt to accommodate this trend.
AMEN.
And there are millions of people who do not have the Google bar installed into their browser.
In this reference, the information from the toolbar is probably being used to determine trends. You don't need everyone to use the toolbar to collect valid data, just enough to form a valid sample.
It is like saying people who consume large quantities of fiber in their food suffer lower incidences of cancer. The effect could have been <i>equally</i> due to their better overall healthy living trend. They may do more exercises, eat better and probably inherit better genes. You get the drift ?
The increase in rankings owing to a correlation in Google analytics is probably due to other reasons. Yes. Google and other major search engines could and do track the click through rates in the SERP and factor this into the rankings. But there is little evidence to suggest that they would be using Analytics data as part of this.
There are millions of sites who don;t use Google Analytics.
Every site wouldn't need to use Analytics. The study I referred to was not testing pages from different sites, and it is unlikely that Google would use analytics data to compare two sites. The theory that was proposed was that Analytics data might be used by Google to determine user behavior on a site and could affect the rankings of different pages on that site relative to each other. For example, if everyone who does a search and goes to your site immediately goes to another page, that other page may replace the original in the rankings or be indented below the original. Toolbar data and SERP click logs are more likely to affect universal rankings.
mjtaylor
12-12-2007, 01:53 PM
Just as wige pointed out above, Google does return a site to you higher if you click on it a couple of times, but that only influences *your* results.
And Google can, indeed, extrapolate to a greater population from the sample of users who have the toolbar installed. There is a question as to whether it is a valid sample, but the information is there and could be used to affect SERPs.
Even without the toolbar or gmail, Google can track whether a site is clicked within each search query and I would bet a pretty penny that information *is* factored into SERPs along with toolbar, gmail and GA information.
NetProwler
12-13-2007, 12:37 AM
Wige - I like your footer - The best way to learn anything, is to question everything.
I agree with mjtaylor on everything except the bit about GA. Wige's footer sums up the reason for it.
As a programmer who once was involved in the development of a search engine, when you factor in X in the weighting and when you get a large body of data without any value for X, you have opened a Pandora's box. The results will be skewed and as you add more such factors, you are unlikely to deliver any repetitive results.
I don't rule out the fact that Google delivers results which vary a lot within a short time span and they factor in hundreds of factors in weighting. Why add one more to the maelstrom ?