Analytics
3K members online now
3K members online now
Learn how to set up goals and Ecommerce tracking, verify accuracy of attribution traffic sources, and discuss funnel visualization, Multi-Channel Funnels, the Goal Flow Report, and Enhanced Ecommerce
 
Guide Me
star_border
Reply

Large abandonment variance between Enhanced E-Commerce and Back-end

Visitor ✭ ✭ ✭
# 1
Visitor ✭ ✭ ✭

I've recently implemented Enhanced E-Commerce via GTM in hopes of improving both our reporting capability AND accuracy.  I've only compiled a couple days worth of data so far (Apr 4 & 5), but there's a big variance in abandonment rates between EEC and our back end, and I'm trying to figure out why.

 

The back end reporting is very simple.  Any cart started and not converted to an order is an abandon.  It's saying 302 abandons for the timeperiod.  GA/EEC is saying 409 abandons for the same period.  That's a pretty big difference! 

 

In terms of orders - GA and our back end are very close.  I can attribute the small difference (6 orders) to timing differences on either end of the timeperiod.  But the abandons is baffling, I expected the numbers to be alot closer.  I am wondering if I am not comparing apples to apples.  But then again, a cart is a cart and an order is an order, there's not a whole lot of gray area here.  Unless I am missing something.

 

I've been through the debugging in preview mode many times, and the events are firing as they should be.  Has anyone faced this issue before, or can anyone shed any light on why there might be such a big gap here?

Large abandonment variance between Enhanced E-Commerce and Back-end

Follower ✭ ☆ ☆
# 2
Follower ✭ ☆ ☆

Hey Rory,

 

You shouldn't compare GA statistics with your backend for metrics that aren't even supposed to align. Your backend has a very different interpretation of an "Abandonment" compared to GA. GA registers an abandonment when the session doesn't have a follow-up step. Sessions can be broken for a multitude of reasons, such as users leaving the site and returning via some other source/medium, waiting 30 minutes before sending another interaction (probably the most common reason in an ecommerce funnel), or due to browsing around midnight of the timezone of the GA profile.

In your data, you can see the relatively thick slices on top of each Shopping Behavior funnel bar as indications of sessions which "started" from that step. Most likely they are sessions which broke in the previous step (e.g. due to the 30 minute timeout) and then the new session that started after the user interacted with the site again is registered as "starting" from the next step in the funnel.

So I wouldn't worry about it. These are not genuine "abandonments" - they're sessions which, for one of a large number of reasons, did not have a complete GA ecommerce flow. The fact that transaction numbers are on par is encouraging, and means that your data should be fine.

 

Simo

Large abandonment variance between Enhanced E-Commerce and Back-end

Visitor ✭ ✭ ✭
# 3
Visitor ✭ ✭ ✭

OK, thanks for responding Simo.  I thought it might be due to session timing, but the session length is configured the same on both ends (30min).  Our site will also register an abandon when a session (with a cart) expires.  So I kind of thought they'd be closer.  In fact, I am kind of surprised GA's numbers are higher - my expectation was 5ish % lower, more similar to orders.

 

Are you aware of what other factors are considered for the "abandoned" bucket on the GA side, other than session length?

Large abandonment variance between Enhanced E-Commerce and Back-end

Follower ✭ ☆ ☆
# 4
Follower ✭ ☆ ☆

Hey Rory,

 

Not session length alone - sessions in general. As I wrote, there are many reasons a session can be broken (with the timeout being by far the most popular one, I guess), which makes trying to consolidate GA's concept of a session with anything home-made quite futile. 

But if you do want to get to the bottom of this, I suggest going about it granularly. Does the number of Add To Cart events recorded in GA match the backend? What about started checkouts?

At what point do you expire the cart? After 30 minutes from adding an item to it? Or 30 minutes from the last interaction of the user? GA breaks the session after 30 minutes of not sending a hit to GA - something that can be quite common especially in ecommerce stores (users taking the last chance to compare competition prices, for example), meaning GA would register the session as a cart abandonment even if the user actually did complete the order, unless you also expire the cart at 30 minutes of inactivity. 

What does "Our site will also register an abandon..." mean - register it in GA or somewhere else?

 

Cart abandonment is simply the number of sessions which had an ADD_TO_CART shopping stage but did not have a CHECKOUT shopping stage.

 

Simo