Google Analytics,
you secretly awesome thang, you.

Three tips for wrangling the beast.



At mosaicHUB, we previously used Google Analytics as a “how are we doing this week?” and Mixpanel as our down-and-dirty. This was all fine and good until we needed answers. These systems were telling us nothing we didn’t already know.

In October, users weren’t spending time on our site like they used to. Last April was a very popular time, but October wasn’t yielding the same enthusiasm from our members. We had made a number of product changes since April which were mostly focused on validating our revenue model. There was no sudden drop in retention nor any complaints from users; there was only a slightly downward trend that persisted for months in a row.

Adding to our blindness was the fact that we implemented Mixpanel in mid-July. It seemed that the entire secret of April’s momentum was trapped in the complicated, clunky box of Google Analytics.

I didn’t like using Google Analytics because it was (is still) complicated. It seems counter-intuitive because I am a full-time web developer, but the truth is that the UI controls are laid out in a way that I couldn’t digest. My attitude towards analytics usually is “the numbers are complex enough… the interface could at least be accommodating”.

I sucked it up and messed around with GA late on a Friday night. By the time I left the office, I was slapping myself in the head for not knowing about the following functionality:


Here are the parts that I discovered that made my least favorite analytics tool 100 times more valuable:



The segment filter

This is how I was finally able to see new-user behavior vs. returning-user behavior (yeah, I am amazed I wasn’t seeing this too.). Mixpanel doesn’t do this as well as I had hoped.


The navigation summary

Comparing the behavior from the mosaicHUB dashboard in April with that in October made it easy to see what was on most user’s priority list. As a heads up, you can use this tool (and every other tool in GA) with a segment filter applied to it.


The acquisition channels

Sometimes, people forget to blame where users come from as a reason for change in site-wide behavior. I knew this existed but I didn’t see the importance until I used it along side the data found with the other two methods of extraction.

We are still working with some of this new data and deciding how to tackle the next big challenges. I’ll save the “how we used this data” section for another time.

I know that some of these GA tips seem trivial. I figured I’d write about it in case you were in the same boat as I was and didn’t bother to dig deeper into my analytics tools until it was too late.

Use these three tips on your own projects and see what you can learn.

Geoff Daigle is the Lead Developer at MosaicHUB.
You can find him on twitter at @dailydaigle