|Analytics: Event tracking vs Virtual pageviews|
Virtual pageviews great in navigation summary
I basically want to know if I can use Event tracking to see what % of people visiting a certain page click to an outgoing link.
I am currently using trackPageview to track outgoing affiliate links and I'm thinking whether I should start using Event tracking instead.
While I understand that virtual pageviews inflate the total number of pageviews (which doesn't bother me as it's not a metric I'm using), I like the fact that by tagging outgoing links as a 'fake page' it then shows in the Navigation Summary.
This not only makes the % Exit more meaningful (I want to make a distinction between people leaving and people exiting by clicking on an aff link) but it helps me find way to increase clickthrough rate for these outgoing links (e.g. see if the aff link is the most clicked link on the page).
Now, I find this quite useful and I'm not sure if I can achieve the same effect using Event tracking. While I can track the aff link event as an advanced segment, I would have to calculate the CTR rate manually whereas if I could get it to show up in Navigation Summary it would show up as a %? Any ideas?
You would have to manually do the math.
On site A we have 2 pages that link to our job section on site B. In event tracking I can segment to the page level and see that site A page 1 had 335 unique events (clicks) and site A page 2 had 110 clicks to the job page on site B. However when I go to navigation summary for site A page 1 or 2 the event clicks are not listed.
So for page 1 I would take 335 and divide it by the unique visitors to get the %.
Checkout the new In-Page Analysis tool in GA, I think that is exactly what you are looking for.
I have very similar needs to hairycoo. I'm currently using Event Tracking for my affiliate links. It's not as flexible as virtual pageviews. I cannot know from which page the user clicked on the affiliate link because Events don't have navigation summary. I will switch to virtual pageviews and create custom reports where I separate real content views from affiliate links...
So, if you're using pageviews, keep doing so. Event tracking -- even though apparently organized -- is pretty useless.
Just to add on what I've learned since I posted this.
By far the biggest benefit of using virtual pageviews over events is the ability to track them as goals (which you can't do with events, or at least you couldn't last time I checked) which opens the door to EXTREMELY actionable and very detailed and useful reports (far more useful than the navigation summary).
It's been invaluable for me even though it's a pain to manually tag every link. Hope this helps.
I wouldn't say event tracking is useless, you have 4 values you can pass, could include the trigger URI for instance.
Current implementation am working on, I've moved away from virtual pageviews towards auto-event tagging. The result is more accurate overall content data for rollup reporting.
The four values you can pass are category, action, label, and an integer value. Each of these values has an intended purpose and none of them is suitable for a "trigger URI".
To me, events are as good as hit counters -- which are not that useful. Events don't have navigation summary and they don't tell whether the user has returned to your site after triggering the event.
On the other hand, if you use virtual pageviews for affiliate links, you have navigation summary, exit percentage, and you can even see the average time a visitor is spending on the vendor's website because Analytics treats the affiliate link click as if it were a pageview of the vendors site!
Implementations don't have to follow the default purpose, its up to you to figure out what works for your setup.
I'm not saying that generating virtual pageviews doesn't have its merits, just in the environment I work in Events makes more sense.
I'm afraid that unless you are doing cross-domain linking (setting up a multi-domain analytics account), you won't get user activity data from another site. Once they leave your site, thats it.
I guess in some cases, like in Google's own video plays example in the documentations, Events can make more sense. But even if Events made more sense than pageviews, if you need more data, go for virtual pageviews.
Also, you can know the average time spent on an external site with virtual pageviews. When a user generates a virtual pageview, _trackPageview is called. Analytics doesn't know the user has left your website. So, once the user comes back to your website and generates another _trackPageview, the difference between the first _trackPageview and the second one is the time the user has spent on the external site. This is also how Analytics calculates average time on page for internal pages -- so it's just as accurate.
Another scenario which I've encountered is an enterprise level site hitting daily maximum unique URIs and dropping 1.5m pageviews over a year (tagged as Other).
Oh, sure, this is assuming that when a user exits your website onto an external site and comes back within half an hour, yes, you can track how long they were elsewhere for.
Still, that's exactly what I was thinking! Making good use of the parameters can help determine what pages the events occurred on.
I just installed the code like this:
_gaq.push(['_trackEvent', 'social-media', 'clicked', 'youtube'])
I'm thinking I may add a suffix to my label, like this:
I could use the action parameter instead for the url "myLandingPage"
not sure which would be the best way...