Virtual pageviews is in fact the way to go with this.  With a virtual pageview setup, your form tag would look something like this:

<form action="x.something" onsubmit="_gaq.push(['_trackPageview', '/submitted-form/']);">

That way, when the form is submitted, it will track a pageview of '/submitted-form/'.  You can then create your goal looking for pageviews of that URL.

The main potential obstacle with this approach is form validation.  If your form uses JavaScript for validation, then you need to integrate the virtual pageview code into the validation code, which can get complicated.  If your form uses server side validation, this method also isn't great - because if the user submits the form and it returns an error, the virtual pageview will still be triggered, incorrectly converting the goal.

</form>

It looks like you've got your ecommerce tracking code incorrectly implemented.  The best way to test this is to run a test transaction, then view source and review what values have been output into the tracking code.  It looks like you are logged a transaction but not actually outputting the relevant prices and other numeric values...

Try applying this Advanced Segment, which will segment out your metrics showing only the traffic which originated from America:

https://www.google.com/analytics/web/template?uid=XGA6--mISZqq4AT60sJBRQ

Once the advanced segment is applied, you can look around on various reports and try to see if you can find anything which seems out of whack.

I was curious about this, so I poked around in the metrics of some of the websites I work on, to see if I could find anything similar (big increase in Firefox (direct) traffic).  I did find that about 80% of all Firefox traffic from mobile devices was showing up as (direct) traffic.

The sites I work on get very little of this type of traffic, so I hadn't really noticed that previously.  In the data I looked at, it didn't really seem like that recent of a change, however.  Maybe you've had a recent surge in mobile traffic?  I think Firefox mobile traffic is typically from Android devices.

There are 2 spots in Google Analytics admin where you enter a URL, the Property Settings and the Profile Settings.  Make sure you've changed to the correct URL in both places.

Could also create two Advanced Segments, one showing traffic only from Safari, the other showing all traffic except from Safari, then compare the metrics to see how the Safari traffic differs from that which is not Safari (assuming it does).

That looks good to me.  If you are unaware, once you get that event tracking properly, you can also setup conversion goals in GA which are triggered by custom events.

You might want to try escaping the question mark, so:

/?specifications

If you want to see the subdomain show up as a referral, you need to remove this line of code from the main domain.

_gaq.push(['_setDomainName', 'mysite.com']);

That should be all you need to change.

The syntax of the GA ecommerce tracking code would be the same on different pages of your site.  When the documentation refers to 'multiple trackers', it's talking about if you want to log the ecommerce transactions into 2 separate GA accounts.

So for your question, same JavaScript code on all receipt pages is what you want.

My opinion would be that DIV-based markup is the better choice here.  As you said yourself, it's not really tabular data, so in using DIVs you can use semantic markup which is a positive for SEO.

You could improve/cleanup the markup of that data though, by:

• Use

  ,

  ,

  tags.  Even the bolded text in the lefthand column are basically headers for the text in the righthand column.

• You should remove the empty class="hr">tags, which I assume are in there to create the horizontal lines.  It's nit picky, as if you remove them, you'll need to add a 'wrapper DIV' surrounding each row, so you won't really be cutting down on the code used that much.  But having empty tags that are only there for presentation purposes is generally frowned upon.  You could create the same visual effect by using a border or by using a background image (if you want the line to not fully extend across the row).

That's all pretty nitpicky coding stuff though.  For SEO purposes, I think the only thing that might have an affect is using the <hx>tags.</hx>

If the partner site would allow you to add code to their thank you page, you could have them just add your existing Google Analytics tracking code to that page and then page views of thankyou.html would start showing up in your Google Analytics.

The problem with that though is that it will show ALL page views of thankyou.html, not just the page views that originate from referrals from your domain.  So this would only work if you are sending traffic to a page on the partner's site that is exclusive to your domain.

You could actually set up a custom report in GA to show only page views of thankyou.html where the referrer is your domain, but you'd have a bunch of additional page views mixed in with your data that have nothing to do with your site.

Actually, site activity generated on other domains than yours will be included in Google Analytics metrics, unless, that is, you filter that activity out.

So you need to create a filter which includes only the traffic from your own domain.  This blog post explains how to do this in a perfect amount of detail.

You might want to try escaping the question mark, so:

/?specifications

I would handle this situation using the URL parameters screen within Google Webmaster Tools (shown in the attached screenshot).  You can add URL parameters into the screen and set them to have 'no effect', which means when Google indexes your site, whatever URL parameters you're set to 'no effect' will be ignored by the spider.

webmaster-tools.png

This is because of Google's (fairly) recent change in reporting of search referrals within Google Analytics.  If the user who visits your site is using Google under SSL (so with a https prefix), their search term is hidden and shows up in Google Analytics as (not provided).  So when you look at the list of search queries in GA, you are only looking at the queries that people searched for when they weren't accessing Google through SSL.  Everything else gets lumped into (not provided).

In Webmaster Tools though, you're able to see ALL of the searches, even the ones from HTTPS searches.  That's what creates the discrepancy.

Don't know if you still need help with this, but here goes...

The main problem is that article you reference is showing out of date Google Analytics tracking code.  The _setDomainName piece that you reference is from the asynchronous version of the tracking code.  The code shown in the article is from the previous version of the GA tracking code.  If you try to mix code from the two different versions, it won't work.

That article is basically just describing how to create a virtual page view that includes the search term the user has searched for.  With the asynchronous code, it would look like:

You would need to update the code above with your own UA ID, domain and replace the 'SearchTerm' placeholder programmatically with whatever the user entered.

Are you sure you definitely want to keep the virtual pageview in there?  I ask because normally you'd have that in place to track clicks on an outbound link, so that you can see how many people clicked it.  But if you're hooking up cross domain tracking to another domain that you control, you can see how many people clicked the link just by checking pageviews of the URL that the link points to.

If you do need to keep it though, your link would look like this:

Checkout Now!

Hi Taysir,

For #1, you can just used the JavaScript GATC.  The vast majority of all smartphones run JavaScript.

For #2, if you are directing all mobile traffic to the mobile subdomain (the m.example.com), and that subdomain is a site built specifically for mobile, I would recommend creating a new Google Analytics property for it (meaning a new UA code).

For #3:

• Make sure you're aware of all the metrics available by default in the Audience > Mobile section of GA.
• It would be worth setting up Advanced Segments for the specific devices that the majority of your mobile users are using.  For example, if you setup an Advanced Segment for iPhone users, you'll be able to see what your bounce rate is just for iPhone users, etc.  If metrics for a specific device are significantly worse than most others, that's a good sign that your site might not be rendering well on that device.

Hi Alfredo,

Your syntax there all looks correct.

I'd recommend though, if it's possible, to instead trigger the event after the form submission has gone through.  In your code above, the event will trigger whether or not the submission goes through, so if the user gets stopped by validation or something, it'll still count as a submission.

In any case, what you have above is correct.