Is CTR playing a role in the Google organic algorithm

[WPsites]

For a good number of years Google's had the code inplace to track clicks on Organic search results.

Data in webmaster tools has confirmed they do store click through rates for clicks on organic search listings.

Do you guys think this CTR is factored into the algorithm? Surely it would be silly not to?

Why not perform a test. As a group we decide on a search term, pick a company thats on the second page, then each day we all click through to the chosen site and see if over say a 2 week period it has any effect its ranking?

[crossdaz]

Why not perform a test. As a group we decide on a search term, pick a company thats on the second page, then each day we all click through to the chosen site and see if over say a 2 week period it has any effect its ranking?

Would that be your site by any chance? :)

Click-through does play a part in rankings.

[directmarketingadvice]

Do you guys think this CTR is factored into the algorithm? Surely it would be silly not to?

No idea, but it would certainly be logical.

Google's big picture goals are relevance and quality. So, "listening" to users to help decide which results are relevent (by using ctr) should help improve the SERPS.

Steve

[Ali-v-8]

it is part of the algorithm that does check ctr and rank you accordingly for that term

[Scott-CopyandDesign]

Wouldn't this make it considerably more difficult for websites to reach the top 1-5 rankings if the existing sites in those positions receive a lot more clicks?

[OldWelshGuy]

What about the 100million other sites that are not on Page1 then? are they doomed forever.

Google have been measuring onmousedown position clicks, but I was told by a google engineer that this is so that they can test the effectiveness of algorithm updates.
e.g if the average click through position for a phrase is 2.3333333, and they introduce an algo change, and the average post change becomes 2.4444444 then the algo have made the serps less relevant as people are having to go further down the page to find relevant info :)

[Ali-v-8]

Its some thing that has been round for a long time.
Google pick up on people by passing the result order.
I dont 100% know how much truth is in this but there are people in the usa and more now in India who get paid to search and click organic listings.

[crossdaz]

What about the 100million other sites that are not on Page1 then? are they doomed forever.


No, because it's only a factor.
If you get a new site indexed where does it first appear - not on a low page for sure? What's been measured in these first few days that can't be measured by any other means?

[OldWelshGuy]

No, because it's only a factor.
If you get a new site indexed where does it first appear - not on a low page for sure? What's been measured in these first few days that can't be measured by any other means?


It doesn't make sense though as the whole system favours first page rankings, unless we are talking about local rank, which is another thing altogether (local rank as in a local elitism algorithm element and not local as in geographic).

By this I mean that the pages in the top 10 might be levated but that would only be temporary. It would also take a heck of a lot of processing power as it is a sub run like hilltop or HITS.

[crossdaz]

It doesn't make sense though as the whole system favours first page rankings,

By this I mean that the pages in the top 10 might be levated but that would only be temporary. It would also take a heck of a lot of processing power as it is a sub run like hilltop or HITS.

This is my theory (and it is a theory, but I've not read anything to disprove it yet?).

Get a new site indexed and it will first appear high in the serps (not necessarily always first page, but certainly where there is traffic). Within a couple of days, and sometimes less, it will drop out of the index entirely and then re-appear some days later at a position which makes more sense.

This is a pattern observed from knocking-out loads of affiliate sites and discussing with people who do the same over the last 2 years.

So what's happening?

My theory is that they want to measure how people react with your site - do they click through and do they bounce back straight back to the serps?
Many successful affiliates will tell you to make sure your site is fully finished and optimised before google finds it because this has a big effect on your sites early performance. I've certainly seen evidence of this at first hand.

So, in my opinion, Google do take account of click through when they find your site and it makes sense that this will be continued afterwards as well. But I'm only 90% sure of it? :)

[OldWelshGuy]

The only problem here is that every single day millions (or more) websites/pages are launched, so how come the Serps are not full of new pages coming and going?

I am not dismissing what you say, just that the eyes show otherwise.

[Scott-CopyandDesign]

The CTR would entirely depend on 4 things:

1). The title
2). The meta desc (or first piece of text on the page if there's no meta desc)
3). The position
4). The visible URL

These are the only visible elements in the SERPs and are therefore the only things which can influence CTR.

All they could do is measure those 4 factors individually (which they do as far as the title tag and URL go), or measure the CTR in relevance to the page and other rankings. This would look at the CTR, but also take into account the average percentage of clicks for that position. So, for example, if a certain position received 5% of all clicks for that search keyword (on average), then this would be taken into account when measuring the CTR. If the CTR percentage is then higher than average, it would show that those four elements are more relevant for that search term compared to other sites in the SERPs (in theory).

This is the only way they could make it fair, as page 2 ranked websites will be getting very few clicks, whilst the top 1-5 will be receiving a lot more.

[crossdaz]

The only problem here is that every single day millions (or more) websites/pages are launched, so how come the Serps are not full of new pages coming and going?


Well if millions of pages are added each day and millions are dropping out every day - how can we know what the total number should be otherwise?

Of course, sods law determines that this won't work but I can get a new site indexed and can you see for your self what happens?
I'm running the risk of looking like a prat but I'm reasonably confident I can replicate what I've observed many times.

When I have chance I'll put a small site on a spare domain and I'll place the first link on here - then we can see.

If nothing else it can be a useful seo experiment for anyone interested?

[Scott-CopyandDesign]

Well if millions of pages are added each day and millions are dropping out every day - how can we know what the total number should be otherwise?

Of course, sods law determines that this won't work but I can get a new site indexed and can you see for your self what happens?
I'm running the risk of looking like a prat but I'm reasonably confident I can replicate what I've observed many times.

When I have chance I'll put a small site on a spare domain and I'll place the first link on here - then we can see.

If nothing else it can be a useful seo experiment for anyone interested?

I've started up at least a dozen websites in my time and none of them have ever magically jumped to page one for keywords I've targeted, or anywhere near.

It could appear near the top, but only for an uncompetitive term I believe.

[crossdaz]

I've started up at least a dozen websites in my time and none of them have ever magically jumped to page one for keywords I've targeted, or anywhere near.

It could appear near the top, but only for an uncompetitive term I believe.

I didn't say that would happen. For a low competition phrase maybe - but for many other terms that would be ridiculous.

I said it will land on a page with traffic. Not page 1000 and not page 1

How close were you watching them?

[Scott-CopyandDesign]

I didn't say that would happen. For a low competition phrase maybe - but for many other terms that would be ridiculous.

I said it will land on a page with traffic. Not page 1000 and not page 1

How close were you watching them?

Very close. I'd know if my current website had been anywhere near page one for any worthwhile terms since it went live.

What page would it land on with traffic? Page 1 has the most traffic, page two as a tiny amount, and the pages after that gain almost no traffic unless someone can't find what they want in the first 2 pages.

[crossdaz]

Very close. I'd know if my current website had been anywhere near page one for any worthwhile terms since it went live.

What page would it land on with traffic? Page 1 has the most traffic, page two as a tiny amount, and the pages after that gain almost no traffic unless someone can't find what they want in the first 2 pages.

Depends what the term is - but certainly a higher position than you'd expect.

Are you saying your twelve sites have each entered the serps at a low position and slowly worked their way up the rankings without ever dropping out of the index? Because if you are then this contradicts everything I know and have witnessed for the last 2 years?

[Scott-CopyandDesign]

Depends what the term is - but certainly a higher position than you'd expect.

Are you saying your twelve sites have each entered the serps at a low position and slowly worked their way up the rankings without ever dropping out of the index? Because if you are then this contradicts everything I know and have witnessed for the last 2 years?

Pretty much.

The competitiveness of the keyword plays a large part in this. You could probably jump up and then drop down for a long-tail keyword. However, if you're targeting a popular keyword, then you aren't going to get anywhere near the top. If I created a site now and focused it on the phrase 'car insurance' it wouldn't be anywhere near page 1.

I'm saying this in relation to your theory. If it does jump up, then I don't think it's because it wants to measure CTR. I simply think Google roughly analyses the website, gives it a ranking, and then makes the ranking more accurate as the Google spider finds more backlinks and begins to properly understand exactly how relevant the site is for a specific search term.

If you're confident that it would work though, then by all means go ahead.

[directmarketingadvice]

This would look at the CTR, but also take into account the average percentage of clicks for that position. So, for example, if a certain position received 5% of all clicks for that search keyword (on average), then this would be taken into account when measuring the CTR. If the CTR percentage is then higher than average, it would show that those four elements are more relevant for that search term compared to other sites in the SERPs (in theory).

They already do that with adwords, so I don't think it'd be much of a problem to also do it with the organic.

Steve

[sirearl]

Always worked on the basis that google does.

I mean I can't see google contacting sites with 10 uniques a day over its adsense.?

Earl

[crossdaz]

If I created a site now and focused it on the phrase 'car insurance' it wouldn't be anywhere near page 1.

I know - otherwise I'd be doing new sites every 10 minutes?


I'm saying this in relation to your theory. If it does jump up, then I don't think it's because it wants to measure CTR. I simply think Google roughly analyses the website, gives it a ranking, and then makes the ranking more accurate as the Google spider finds more backlinks and begins to properly understand exactly how relevant the site is for a specific search term.

This could be possible although it doesn't explain why the site then drops from the index entirely?

As has been mentioned it's an established fact with adwords so it's reasonable to say that natural results will be the same?
However, I can't prove this - only work on the assumption it does. It certainly can't do any harm to go with this theory as click through is what you want at the end of the day.

[OldWelshGuy]

I think maybe there was a bit of misunderstanding in what I was saying. What I was saying was that the 'jumping in' thing (that everyone knows does often happen and always has) only applies to very low competition phrases on well optimised new sites.

My point was that Google only use things algorithmically that can be applied across the board. Remember that Google are struggling with processing resources on a costant basis.

[Birmingham]

Yes of course as you say it would be silly not to, so even if they don't now they will do when they wise up a bit.

For a good number of years Google's had the code inplace to track clicks on Organic search results.

Do you guys think this CTR is factored into the algorithm? Surely it would be silly not to?