We use cookies. You have options. Cookies help us keep the site running smoothly and inform some of our advertising, but if you’d like to make adjustments, you can visit our Cookie Notice page for more information.
We’d like to use cookies on your device. Cookies help us keep the site running smoothly and inform some of our advertising, but how we use them is entirely up to you. Accept our recommended settings or customise them to your wishes.
×

Google Analytics Funnels: How to interpret the data and avoid one big misunderstanding!

Google Analytics Funnel Visualisation report may work in a very different way to your expectations. Read more to understand what you need to be looking for when analysing your funnels, and how you may have been coming to incorrect conclusions before!

The background

I won't go into too much detail regarding the background here, save for:

  1. In Google Analytics (GA), it is possible to understand the proportion of visitors that complete a given process on your website, by adding a funnel to your goal, in the Goal creation view. One point I will stress here is that the function of these reports is to help you get insight into how your structured process performs. Note: structured process. So we're talking about checkouts, donation process or multi-form account sign-ups - we're not talking about trying to use a funnel to understand what proportion of your visitors landed on the homepage, went to make a cup of tea, got a haircut then came back to the site and went straight on to purchase.
  2. In Google Analytics there are two ways of visualising this data; the Funnel Visualisationand Goal Flow reports. Despite appearing to simply be two different ways of viewing the same data, there's actually a multitude of differences between the two channels, and the data is not entirely comparable. Handily, Google have very eloquently documented these differences here.

So what's the problem?

We consider ourselves to be pretty seasoned veterans at the ol' Google Analytics, but even we were completely thrown by this one. After some analysis and questioning of GA, we learnt something which fundamentally changed what we believed to be true about Google Analytics funnels and the Funnel Visualisation report.

Namely, if a visitor 'exits' your funnel (either by going back a step, refreshing the page they are on or by viewing a page outside the funnel process, e.g. by opening terms and conditions) at any step, it means that they definitely didn't complete the process in that session.

In other words, whereas previously you might have observed a large drop-off from a given point in your funnel and seen lots of traffic going back from, say, Payment Details to Address Details and thought 'huh, lots of people need to go back and change their address before going on to complete their purchase, how odd' - this isn't the case. These visitors did not complete the purchase in that visit.

Example

Examples help, so please see below for an example funnel from an e-commerce retailer. Let's focus on before/after analysis of their funnel now we know what we do.

Funnel Visualisation report

Before: without our new knowledge, we would have looked at this funnel, seen an almost 30% drop-off from Shipping Details to Payment Method and thought, 'that needs investigating'. Closer analysis reveals that only 118 of the 497 vistors actually left the site (indicated by the row saying (exit) - so what if a bunch of customers needed to go back to Cart or Address Details, there's not a problem here.

After: Again we would have looked at the funnel, but now we understand that we lost ALL the traffic on the right hand side. None of those 497 went on to complete their order despite re-entering at a previous step in the process. So, this requires more serious investigation than previously thought.

Technical background

In a nutshell, a given visit can only have one entrance to a funnel, and one outcome from that funnel. If there are multiple entrances, GA takes the earliest possible step as the entrance (note: not the latest one) and if there are multiple exits, GA takes the step furthest into the process. So if a visitor does the below:

Step 1 > Step 2 > Step 3 > Step 2 > Step 3 > Step 4 > Step 5 (completion)

GA reports the below:

  1. Visitor enters at step 1
  2. Visitor exits at step 3 (to go back to step 2)
  3. The entrance at step 2 is ignored as it's further along than step 1
  4. On completion of the process, that completion is taken as the furthest outcome and so point 2 (the exit from step 3) is removed.

What you should now see then is that wherever there was an exit, the visitor cannot have returned to complete the process.

Equally, you would never see the loopback in the Funnel Visualisation report. This is a linear view of the proportion of traffic who saw each step in your process. To see loopbacks and skipping of steps, Goal Flow is your friend, as you can see below.

Goal Flow report

What next?

Revisit your funnel analysis! Look out for the dreaded red and take steps to increase the proportion of your traffic that moves from step to step. Remember, testing your changes means you're more likely to improve the performance of your site than if you just guess!

Join the Discussion