Double Meta Refresh - a few questions

Status
Not open for further replies.

Cataclysmic

New member
Jan 31, 2009
164
2
0
Canada
1) When using a Double Meta Refresh, what is it about the process that blanks the referrer? Is it simply the fact that you're doing a refresh - does a refresh always blank the referrer? (I can't see anything else in the code that means "blank the referrer")

2) The way I understand it, a single meta refresh would actually blank the referrer most of the time. The reason to do a double meta refresh is just to make certain that the referrer is successfully blanked, and if it isn't, to take the user somewhere other than your affiliate offer page. Is that correct?

3) When you do a DMR, your traffic is going like this:
Website.com >> Example.com/redirect1.php >> Example.com/redirect2.php >> AffiliateOffer.com​
Right?
When this happens AffiliateOffer.com does not know that you originally came from Website.com. Does it look to AffiliateOffer.com like you only came from Example.com/redirect2.php, or does AffiliateOffer.com have no idea whatsoever where you came from?

4) If you redirect your traffic through the same path as shown in #3 above, but instead of doing meta refreshes you use regular php redirects like this...
<? header("location:http://AffiliateOffer.com"); ?>​
...will AffiliateOffer.com know that you originally came from Website.com? If yes, how would they know - does the entire series of URLs that you passed through get sent to AffiliateOffer.com?

5) If I use a DMR to link my landing page to the affiliate offer, does this somehow prevent the GoogleBot from knowing that the LP links to the affiliate offer? If no, is there any other way to prevent the GoogleBot from knowing this (thereby resulting in a higher QS for ads/keywords that link to that landing page)?

Thanks in advance for answers.
 


Ok, thanks. That's a partial answer to one of my questions, but since you don't sound confident about it, I'm still looking for a complete and definite answer, as well as answers for the rest of the questions...
 
Yeah I read that explanation too and it's great... but it was AFTER reading that explanation, and a few other ones, that I wrote my original post, because those questions are not answered.
 
1) DMR clears all referer headers, it masks the source of the traffic (ie your lp) by blanking the referer section of the HTTP headers that the browser sends

As for "why", by the looks of the code you posted he doesn't know why either. My best guess is it just confuses the browser so it clears the referrer in the same way a typed-in address has no referer.

2) Yes correct, the second refresh is to check if the header has cleared.

3) Referer header cleared means there is no record of the source of the click.

4) AffiliateOffer.Com never knows about Website.com, unless the link from website.com goes straight to AffiliateOffer.com or you pass them the source of the link

5) Can't answer definitively, googlebot is pretty clever. There are some javascript tricks I can think of off the top of my head that will hide the links from non-js processing bots.
 
The php redirect passes the referrer, if it's there to pass... end of story, unless you further manipulate it down the track.
As for the meta or html redirect, this is what I've read, but I doubt you'll ever get a definitive answer -if a meta or redirect is only of short duration, Google will follow the redirection & index the final destination page only. But, if the meta is 3 sec or longer then Google will index both pages. YMMV
 
Thanks. Rep given. A few further clarifications...

Cataclysmic said:
4) If you redirect your traffic through the same path as shown in #3 above, but instead of doing meta refreshes you use regular php redirects like this...
<? header("location:http://AffiliateOffer.com"); ?>
...will AffiliateOffer.com know that you originally came from Website.com? If yes, how would they know - does the entire series of URLs that you passed through get sent to AffiliateOffer.com?

4) AffiliateOffer.Com never knows about Website.com, unless the link from website.com goes straight to AffiliateOffer.com or you pass them the source of the link

My experience does not match this, because I had a situation where I was sending traffic through a php redirect and somehow the advertiser knew the original source of my traffic.

But assuming my experience was some sort of fluke and that what you wrote is true... let's say that I'm advertising on the Google Content Network. I don't want the affiliate network or advertiser to know which websites my GCN ads got clicked on, which converted, etc.

So if I send my traffic through a php redirect like this:

GCN Website >> MyRandomWebsite.com/redirect.php >> AffiliateOffer.com

...you're saying that AffiliateOffer.com will see that the traffic comes from MyRandomWebsite.com/redirect.php but will know nothing about the GCN Website, right?

It seems to me that this is just as effective as a DMR because I really don't care if they see that the traffic came from MyRandomWebsite.com/redirect.php. That knowledge is useless to them, isn't it? They can't use it to see how I'm advertising and compete with me.

Am I understanding all this correctly?
 
...
So if I send my traffic through a php redirect like this:

GCN Website >> MyRandomWebsite.com/redirect.php >> AffiliateOffer.com

...you're saying that AffiliateOffer.com will see that the traffic comes from MyRandomWebsite.com/redirect.php but will know nothing about the GCN Website, right?

It seems to me that this is just as effective as a DMR because I really don't care if they see that the traffic came from MyRandomWebsite.com/redirect.php. That knowledge is useless to them, isn't it? They can't use it to see how I'm advertising and compete with me.

Am I understanding all this correctly?

As has been stated before, a php redirect carries all the information of original referrer. So in the case you listed above, Affiliateoffer.com will see that GCN sent the visitor. Test this yourself using 2 of your sites. A link on site 1 that points to site 2 that has a redirect back to a page on site 1 that captures the referrer.
 
As has been stated before, a php redirect carries all the information of original referrer. So in the case you listed above, Affiliateoffer.com will see that GCN sent the visitor. Test this yourself using 2 of your sites. A link on site 1 that points to site 2 that has a redirect back to a page on site 1 that captures the referrer.

Thanks for the information. +rep for you too. What you have written is consistent with my experience, but contrary to what Metcalfe wrote above - "AffiliateOffer.Com never knows about Website.com".

I'd test this if I knew how... I can implement the redirects no problem, but I don't know how to tell which referrers get passed along or not. Is this simple? If yes, can you explain how to check it? Thanks.
 
Thanks for the information. +rep for you too. What you have written is consistent with my experience, but contrary to what Metcalfe wrote above - "AffiliateOffer.Com never knows about Website.com".

I'd test this if I knew how... I can implement the redirects no problem, but I don't know how to tell which referrers get passed along or not. Is this simple? If yes, can you explain how to check it? Thanks.

The output below spits out the referer if one was passed.
echo $_SERVER['HTTP_REFERER'];
 
Let's break it down to the simplest of possible examples, I just walked through this and it satisfied my curiosity

Setup 4 files on your server:


redirecttest1.html

Code:
<a href="http://YOURSITEADDRESS/redirecttest2.php">TEST (blanked referrer)</a><br>
<a href="http://YOURSITEADDRESS/redirecttest3.php">TEST (intact referrer)</a><br>
redirecttest2.php
Code:
<?php
echo "<meta http-equiv=\"refresh\" content=\"0;url=http://YOURSITEADDRESS/redirecttest3.php\">";
?>
redirecttest3.php
Code:
   <?php
$referer = $_SERVER['HTTP_REFERER'];
if($referer == "")
    {
echo "<meta http-equiv=\"refresh\" content=\"0;url=http://YOURSITEADDRESS/redirect4.php\">";    
    }
else
    {
echo "Referrer: ".$referer;
    }
?>
redirecttest4.php
Code:
<?php
echo "Referrer: ".$_SERVER['HTTP_REFERER'];
?>
Click back and forth on the two links in the first file and you will see how it works.

But here is the summary:

Clicking on "TEST (blanked referrer)" takes us to redirecttest4.php and we see a screen that says "Referrer: "

Clicking on "TEST (intact referrer)" takes us to redirecttest3.php and we see a screen that says "Referrer: http://YOURSITEADDRESS/redirecttest1.html"



My experience does not match this
...
It seems to me that this is just as effective as a DMR because I really don't care if they see that the traffic came from MyRandomWebsite.com/redirect.php. That knowledge is useless to them, isn't it? They can't use it to see how I'm advertising and compete with me.

Am I understanding all this correctly?

It's not useless, Knowing your web address is enough for them to find out where your ads are running if they know some tricks.

The other part of this is that some browsers hold onto the HTTP_REFERER header through a meta refresh, and if you just send them off to the next site they will know your original traffic source.

It is insurance against that.
 
With double meta they won't have a clue where the visitor came from. PHP header redirects carry the referer, otherwise I'd be very rich right now :)
 
Metcalfe - Thanks for posting the code you used for your test. I just used it to run a more detailed test. Whereas you tested out 2 different paths to the final page (one via double meta refresh, and one not), I tested out 6 different paths....

1) Regular link direct to final destination page - referrer got passed (obviously)

2) Single PHP redirect to final destination page - original referrer got passed

3) Double PHP redirect to final destination page - original referrer got passed

4) Double Meta Refresh - referrer was blank*

5) Prosper202 non-cloaked link - original referrer got passed

6) Prosper202 cloaked link - referrer was NOT blank - but what got passed what not the original referrer; it was the URL where I have P202 installed.*​

*When using Firefox, the DMR worked fine. When using Google Chrome, the DMR did NOT work - the referrer did not get blanked (I guess Chrome is one of the browsers that it doesn't work on - not good since Chrome is getting popular). The P202 cloaked link (path #6) worked the same in both Chrome and Firefox.

I think I'll be using P202 for my cloaking, since it seems to work consistently whereas the DMR does not. I'd prefer if my P202 URL didn't get passed through, but I can't think of any way that can actually hurt me.... can it?
 
Status
Not open for further replies.