BigCommerce Analytics vs. Google Analytics
Thought I’d shed some light on these 2 different reporting platforms as we’ve been experiencing tremendous sales growth and it is important to us as a company to determine the accuracy and efficiency of our marketing methods as well as identifying the reasons for our growth. unfortunately, we’ve noticed these platforms reporting numbers that were extremely off from one another and have been resilient in finding out why!
In the beginning, we set up Google Analytics just as described in Big Commerce (Aug 2012). Since then, reported sales / visitors / conversions have been inaccurate reporting numbers as low as 60% of what BC was reporting. This is frustrating, because as you know, BC cannot do a lot of the great features GA can, like exporting data, or reporting pageviews, goals, conversions etc!
Immediately set up SSL for goal conversion accuracy!
As part of my job in marketing, I use email marketing as a way to direct traffic to the store. Well, it’d be nice to know how much of this traffic is converting, but every time I set up a Google tagged URL (http://support.google.com/analytics/bin/…r=1033867) to track campaigns, we saw tons of people visiting, but nobody converting. I thought this was odd, so I attempted to use my own custom URL, buy a product and see what happened. As expected, I saw my traffic but no conversion. After I set up a 3rd party SSL, this started working. In short, the shared SSL that BC stores use by default confuses GA as it switches your stores root directory (xxx.mybigcommerce.com) at checkout, therefore making goal conversions ineffective.
This was a surprising find for us – Once we got goal conversions working, we still noticed order and order size discrepancies. It seemed orders that BC were reporting weren’t being reported in GA until the next day. So who’s right? After confirming both platforms were set to our timezone (EST) we simply looked at the timestamp on an order email confirmation. BC seemed to think it was placed an hour prior, meaning our EST setting was having no effect. Thus, GA was correct! This a bug BC, if you’re going to advise us to set our timezone, at least make it work!
EDIT – This is because Daylight Savings was still checked under store settings. Oops! Easily overlooked though!
What is ‘an order’?
This was an interesting experiment – in comparing number of orders from BC to GA, the number of orders were different. Additionally, in analyzing an number of orders from a particular day on BC and then looking at that SAME number of orders say, 2 weeks later on BC, we would notice the number drop 1 or 2 orders. Why?
- Refunding (Even partial refunds!) warrants BC to subtract that order from the order number total.
- GA doesn’t care about refunds.
- BC will report orders placed on back end.
- GA will not.
So thats a tricky number to look at and it’s accuracy will depend on how you deal with refunds, etc. We’re thinking it may make more sense to start utilizing store credit as it will push any corrected orders customers want to the goal page that GA can utilize, therefore taking back end BC orders out of the equation.
There’s still tons more to figure out. The next thing we’re going to look into is Unique Visitors, as GA and BC both seem confused as to what this metric truly means. In the mean time, I hope this post sheds some light on these sometimes confusing platforms and also pushes BC to step up their development from a reporting standpoint.