Google Analytics setup tips and tricks?

[HashtagJeff]

I want to see (and possibly compile) other marketers tips and tricks when it came to setting up their Google Analytics. I feel like I sometimes get into a rut when I setup GA for clients and I want to make sure that I'm not getting rusty with it.

Some of the things I'm interested in (but not limited to):

• How do you use dashboards?
• What funnels do you setup?
• How do you utilize goals?
• What are your favorite metrics to report to clients?
• What are the top items on your checklist when setting up GA?
• And many others I could include here...

I'd be interested in seeing what other marketers include in their GA setup and reporting. These can be super simple or expert level tips and tricks. It's a personal curiosity I have so that I make sure I'm staying up to date with everything but if there is enough participation I will compile everything into a YouMoz post or something as a guide for us to use and share.

So please share this question with others if you can. Thanks!

[ecommercebc]

Hey there,

Thought I'd get the discussion started...Hopefully! I'm using tag manager more and more. I'm really enjoying the debugging interface and the layers you can apply onto events/actions. I've yet to use custom variables in tag manager though, so would be keen to hear about how people have got on with this.

I'm a big fan of goals and solid ecomms tracking. However, one thing that I have noticed which causes a bit of tension is event tracking labelling. For action, I often put the verb, e.g 'played/watched/clicked' and for label the filename/video name etc. However, some GA event tracking plugins and documentation track the filename as the action and label as the verb. I'd be keen to hear if anyone has a preference one way or the other! It feels kind of wrong to put the label as the verb, but I'd definitely change my mind if there was a compelling argument.

[ColeLusby]

I use dashboards for weekly / monthly reports showing basic data such as: users, bounce rate, locations, conversions, avg time spent on site, top pages, browsers, device, etc.

My favorite trick is the (Not Provided) filter I implement under Acquisition -> Channels -> Organic Search. So, people can see what pages (which you can determine keywords from that page) are bringing in the most traffic.

I like implementing a small code within the GA script to say "if user spends x amount of time on page, then it doesn't count as a bounce rate" - because the goal of the bounce rate is to show who isn't engaged. But if you visit one page, you could be highly engaged and leave.

Those are my favorites!

[Martijn_Scheijbeler]

Hi Cole,

Curious how you change the GA script to not add people to the bounces 'segment'. Why not chose to use event tracking on these interactions on the page so they won't be counted as a bounce?

[ColeLusby]

Hey Martijn,

I suppose that would be an alternative - I've never counted someone as an event personally.

However, the goal of this is to have an accurate percentage of people that are not engaged. This would be applied to all pages (assuming you have many pages or articles). I'm not familiar if the events "tip" would be applied to all pages. Can you explain the event tracking on specific interactions? Keep in mind this is for all pages / posts / articles, manually adding event tracking (if that's the case) is inefficient.

Here is a link to an article that I found helpful: http://www.hmtweb.com/marketing-blog/adjusted-bounce-rate-in-google-analytics/

Thanks,

Cole

[ecommercebc]

Hey Cole,

You can definitely apply event tracking on an automated basis across an entire website. This is something I've done several times before. I guess it depends on what event you'd track?

You can implement it for things such as scroll to understand how much of your content users are getting through, which might be relevant for you: http://cutroni.com/blog/2012/02/21/advanced-content-tracking-with-google-analytics-part-1/

I think for your goal though, I can't think of an obvious implementation that would interact with bounce rate like your script adjustment does. I guess it depends on your preference of tracking - for me I'd probably opt to use some form of event tracking, like Martijn.

[Martijn_Scheijbeler]

Hi Cole,

Yeah well that change in script is in close proximity to what you would normally do with adding event tracking to real interactions on a page and in this case you 'fake' an interaction after 15 or X seconds. I still wouldn't prefer using it this way though as you don't know for sure that the user is actually in the browser window or at all looking at the page, so I would at least check for that with some additional JS but in the end you'll get the same result.