But Shawn, it's not just about branded keywords, is it? Everything that lands on that page has nooveride applied to it. So no matter how someone gets to that page - organic, referral, a new click on a coded link -- the first source (or more correctly, the last source they used before you applied the nooveride parameter, subject to GA's pecking order) gets credit.

Wouldn't it be more appropriate in this special case (branded keywords) to apply &utm_nooveride=1 to the link in the ad, so that only the branded ad never gets credit unless it is the first contact the visitor had with the site?

@Robbin:

Your first paragraph is a little off base. My article does not suggest doing things in this fashion, and here's why:

'Everything that lands on that page has nooveride applied to it.'

This is incorrect. My article only suggests modifying the destination URLs of branded AdWords ads. In addition, what you suggest is actually impossible.

Secondly, your second paragraph basically summarized my article. I specifically state that this only applies to branded campaigns for cpc, as you can use this method for Yahoo and MSN as well.

I guess my point is that I don't know where you got the idea that I was talking about all sources. My article is only talking about branded cpc campaigns and changing the destination URLs of the ads.

We do this for internal email promotions as well.

Basically, we try to add utm_nooverride to any non paid incoming link so we can get a better handle on the true cost per sale.

On a related note, I've been looking into solutions for GA that will track ALL touch-points (campaigns) for a sale, but haven't found anything I really liked.

@Ophir:

Stay tuned for Parts 2 and 3. 2 will cover when and when not to use this method for emails, and part 3 will focus on special situations.

Hi Shawn. From the way you wrote it (or you could say, from the way I read it!!) it felt like you were going to hard code the page that way. So the URL of the landing page (not of the destination uri in AW or Yahoo or MSN) would be www.site.com/landingpage?utm_nooveride=1

Well, I guess I need to read more carefully next time.

@Robbin:

I suppose you could also say that I need to write more carefully next time. I will admit I'm a little confused as to where the hard-coding idea came from, but I'll be more careful to make my instructions absolutely clear. Thanks for the feedback!

Shawn. Please could you clarify. Lets say the current destination URL in our YSM "Brand" ad group is -http://www.oranges.com/index.htm?utm_campaign=
utm_source=ysm&utm_medium=cpc.
Do we need to add utm_nooverride=1 to the beginning/end of the above tagged url or do we change the entire url to -http://www.oranges.com/index.htm?utm_nooverride=1?

If the latter, what happens if a visitor has not clicked on a different ad before and there is nothing to override. Presumably GA then doesn't know that this is a YSM-CPC visit at all?

Also, can we apply this with AdWords autotagging on PPC ads?

@Nick:

Great Questions.

First, you don't replace the manual tags for Yahoo! and MSN campaigns, simply add the utm_nooverride=1 parameter to the end of the query string. Also, there is an error in your example's URL and you've left out the keyword info!

So the new Branded destination URL will look something like:

www.oranges.com/index.htm?utm_campaign=Branded
&utm_source=yahoo&utm_medium=cpc
&utm_term=oranges.com&utm_nooverride=1

Secondly, the way utm_nooverride works is that cookies are not overwritten only if they already exist. If there are no cookies, it will set them using the manual tags like it normally would.

And finally, yes, you can do this with AdWords ads.

Hope you found this helpful.

I have been curious about this very issue, thank you for jumping into it. Will this technique work for Urchin 5.6 as well?

@Karl: Yessir. It ought to work with the UTM version 4 and up, I believe.

Another great article Shawn!

I was wondering though how this variable might interact the gclid if you use the autotagging functionality in AdWords?

My hunch is that it ought to work just the same as what you describe with manual tagging, but I was wondering if you'd tested it and could confirm for sure.

Thanks for you time.

@Alan:

Thanks! The utm_nooverride variable works fine with autotagging in AdWords. This was actually my motivation for writing the article in the first place, since it's so easy to set up, and some of our clients have found it to be very useful.

Thanks for reading!

Shawn. I am a little late to the party, and I have a unique situation. We get a lot of traffic from organic traffic. We get traffic from a variety of sources, and they browse our site a lot, then they come back like 2 weeks later by searching for "fiftyflowers" in Google for example. They click on our link, which may be organic or PPC, and buy something. I want to know the original source of this conversion. Can I use the utm_nooverride paramters? I am especially interested in the affect it would have on the organic traffic we get.

Thanks in advance!

Hi Blu,

Great question. Currently, we've confirmed that you can use the utm_nooverride parameter for your PPC ads (that's exactly what this article is all about). This works for both Google and non-Google PPC ads.

As for organic, I have yet to see a great way to do this, as you can't directly change the destination URL's in your search result entries. There are a few special variables that Google Analytics uses to count certain keywords as direct traffic, but there really hasn't been a lot of documentation out there on how to use them, and we haven't yet had a chance to test them.

You can find the versions of these variables for both urchin.js users and the newer ga.js by looking at Google's Migration Guide (which was the first I'd heard of this). Look for the section called Setting keyword ignore preferences. If we can find time to test this and get any substantial results, we'll be sure to blog about it.

Thanks for reading!

Hi Shawn,
congratulation for your article, great informations for a beginner like me in the world of analytics !

Like Blu, we encounter a problem with returning visitors from Organic traffic.
Our problem :
They find us from our ads, visit our website, go on some others websites, and finally come back to us and proceed booking after typing the name of our website in google to find a link to our website...

I was thinking, maybe it's possible to use different links for organic and Ads.
By using an url with the "nooverride" variable set for organic (the main index website url) and a second url for ads without that variable set.

What do you think ?
Maybe using the nooverride by default is not a good idea ?

Thanks !

@Yan:

The problem with most organic traffic is that you usually have limited control over which URLs show up and which 'ads' are displayed. Even if you are focusing on SEO, the secret formulas that search engines use are constantly changing, making you have even less control over what happens.

Because of the limited control, it has been difficult to find methods using utm_nooverride for people who want to 'ignore' certain kinds of organic traffic. I would say that for the most part, leaving your organic traffic alone is just fine. I think it's important to understand how exactly users are finding your site, and if they are returning via organic branded terms, then you would probably want to know that - especially if you spend a lot on SEO.

I think using utm_nooverride in a blanket fashion is not a good idea personally. The web is a very dynamic place, and I think it's important to measure the way users find and interact with your site in the most 'real' way possible. So if users are clicking on your AdWords ads and then later finding you via a branded organic search, you can measure it!

That's not to say there aren't times when you want to keep the original source. I just think that organic traffic should be measured the normal Google Analytics way, without any kind of utm_nooverride manipulation.

On top of all of that, I think implementing this method would be difficult to do (if not impossible), and would most likely have some unexpected negatives that would outweigh the benefits.

Thanks for reading and for bringing up such an interesting topic!

Very interesting indeed! Seperating brand name conversion is also possible in AdWords if you were to set up a different account for the branding campaign. This way, your initial conversion will not be overwritten by a conversion on brand name. You will, however, have double counts but I wouldn't know a way to measure how many (which part of the conversions in the branding account really converted on your brand name and how many first visited your site via another keyword and is thus registered twice: once in your regular account and once in your branding account). Obviously, this is not an ideal method. Would anyone know other related articles on conversions on branding campaigns and how to analyse/decide which percentage of those conversion actually converted on brand name and which percentage first find your website via other keywords and then converted?

@Rene:

I don't think I'd use the additional AdWords account for this - it's best to have everything under one account if possible.

As to the question of finding out how to better analyze the initial source of your brand name traffic - I would say that you can be creative with cookies on your website to store this information. You can always create cookies designed to measure a user's initial source if have experience with them.

Other than that - I recommend looking at correlations with other types of traffic, and also using simpler methods such as utm_nooverride. Essentially, this method lets you see most recent conversion sources (regardless of brand) within AdWords, but within Google Analytics these repeat branded users will be reallocated to their previous sources.

Shawn,

You stated that AdWords conversion tracker already contributes the conversion to the latest keyword used, correct? Google states the opposite: "if a user clicks on a keywords twive, the conversion is attributed to the last keyword". Do you disagree or did I not interprete your words correctly? Thanks!

@ Rene: Maybe I'm not reading your comment correctly, but it seems you suggest Google and I are actually saying the same thing, not the opposite. Regardless, both are correct. AdWords Conversion Tracker attributes a conversion to the last/latest AdWords ad/keyword that was clicked (in the last 30 days). In other words, if I click on the same keyword twice, the second one will get the credit.

Shawn, sorry, my mistake. You Said: "this [the no-override way] is how Google AdWords Conversion Tracker already tracks things. [...] So this information should already be readily accessible."
However, AdWords is NOT measuring in the same way as the no-override: with the no-override the first keyword gets the credit, in adwords it will be the last. I misquoted you earlier, sorry. Do you agree on this?

@Rene:

Looking back at that part of the entry, I'm honestly not 100% sure what I was trying to say. I can only assume I meant something else by the word 'this'.

I think what I meant was this: AdWords works very much like the utm_nooverride system in Google Analytics when it comes to visitors that return from other sources. In other words, if I click on an AdWords ad on Tuesday, but come back from a non-AdWords source and convert on Saturday, AdWords will still properly track the conversion.

However, AdWords does not track this way when it comes to return AdWords ads, as you suggest. Which is really the whole reason I wrote this article :) Anyway, that's definitely an error on my part, and I will update the post to try and clarify this. Thanks for reading!