Before we look at the full customer journey, let’s zero in on the website funnel to isolate the onsite user experience and improve conversion rates.
Say your website conversion rate is <2% and you want to increase that. Where do you start?
Firstly, you should acknowledge the inconvenient truth: your website doesn’t have one funnel:
The typical path to purchase for a typical ecommerce website
Your website has multiple funnels:
The above diagram is a simplification.The point is, throughout your funnel, website visitors are abandoning the process. They may drop off at any stage.
The goal of funnel analysis is to identify where are you leaking visitors and to what degree.
Example #1: “on step 2 of the checkout process, we lost 80% of our iPhone visitors, compared to 50% of our Android visitors. Something is wrong with our iPhone experience on this page”.
Example #2: “the add-to-cart rate on the product page is 10% on desktop and 2% on mobile. We should focus on improving the mobile experience on this page.”
Different funnels track different KPI’s. For example, the primary metric for your newsletter funnel is email opt-ins:
A simple newsletter funnel
Every website is unique, but here are some typical funnel steps for an ecommerce store:
|Product List Views
|Product Page Views
|Add to cart
|Checkout step 1
|Checkout step 2
A segment is a subset of your Analytics data. Segments let you isolate and analyse subsets of data so you can examine them.
In Google Analytics, you should segment your funnel analysis to look at the most popular subsets of users. This makes your funnel analysis more actionable. It’s likely that many conversion bottlenecks are segment-specific (i.e. they don’t apply to EVERYONE).
Segments can take into account any dimension in Google Analytics. Below I have listed the most actionable:
|User type i.e. new vs returning
||Operating System e.g. iOS, Android
||Windows desktop vs Mac desktop
|Device i.e. mobile vs desktop vs tablet
||Iphone Model e.g. iPhone 8, iPhone Xs Max
||Browser e.g. Chrome, IE, Safari
||Android model e.g. Samsung 8, Pixel 2
||Browser versions e.g. Chrome 78, IE12,
Your funnel will isolate these different segments at each stage of the user journey.
For funnel analysis, I highly recommend creating segments with a user scope as opposed to a session scope. Analysing users rather than sessions is better. It’s a more accurate reflection of reality (since one user can generate multiple sessions).
To build some of these segments in Google Analytics, here is a cheatsheet. This document is from my Checkout Optimisation Portal (free to access). There are many, many more segments that can be built.
Clearly, analysing EVERY available segment in your funnel analysis is unrealistic. Apply 80/20 and be practical. Which segments represent a significant proportion of traffic?