Note: This post has moved from Leapthree.com to Ayima as part of the 2018 acquisition.
By now everyone should be excited about the release of Calculated Metrics within Google Analytics. This has the potential to be yet another powerful tool for analysis, although it will only prove useful to companies who are investing the time and resources in a good GA set-up. As a bonus, it shuts down another line in the arguments between Google & Adobe Analytics.
Details on how to create Calculated Metrics can be found in some great blog posts, notably those by LunaMetrics and AnalyticsPros, including a solid list of suggestions to get you started. It is a staggered roll out of the new feature so don’t be alarmed if you don’t have access yet, it shouldn’t be far away from appearing in your GA Account.
But I was confused by these and other blog posts as they appeared to be missing the most obvious and powerful use of Calculated Metrics. Most businesses have some form of funnel at the core of their website. In nearly every GA set-up that L3 Analytics performs, we create a goal for each stage of this process. Our clients can then create a horizontal funnel, with this being an incredibly useful tool for analysing performance.
With Calculated Metrics, you can now create the completion rate between each stage of the funnel. It is as simple as Goal Y Completions / Goal X Completions. This set of calculated metrics can then be used with any session or user based dimension to see where visitors are dropping out of the process. We have been doing this within Excel for years and it is great to finally be able to do it directly within GA. It will speed up the analysis process immensely and offer more flexibility in which dimensions to drill into.
Step by Step Instructions
Step 1 – Create a Goal for each stage of the funnel
As mentioned, most websites have a funnel as the core component of their customer journey. It is obvious for any ecommerce website but also true for booking engines and lead generation websites. As a first step, identify each stage in the funnel, ensure it is being tracked and create a Goal based on the page name or event being fired.
The following set of goals reflects the funnel for a retail website (where the visitor is not taken directly to the basket after creating it). Note that an Ecommerce Session is one where a visitor is interested in a purchase.
Step 2 – Create the Calculated Metrics
The next step is to create a Calculated Metric for the completion rate between each stage of the process. This uses Goal Completions. So the calculations are:
- {{View Product (Goal 2 Completions)}} / {{Ecommerce Session (Goal 1 Completions)}}
- {{Create Basket (Goal 3 Completions)}} / {{View Product (Goal 2 Completions)}}
- {{View Basket (Goal 4 Completions)}} / {{Create Basket (Goal 3 Completions)}}
- And so on…
The formatting type needs to be percent as per above. I discovered that as long as you are creating good Calculated Metric names, the external names will take care of themselves.
Step 3 – Use these Calculated Metrics within Custom Reports
All of these Calculated Metrics can then be used within a custom report. In this example, we will be creating a Funnel metric group. Start the sequence with “Sessions” and “Goal X Conversion Rate” to show total traffic and % of sessions that progress to stage 1 of the funnel. Then list the calculated metrics for completing each stage of the funnel process, finishing with the number of total conversions.
Multiple Metric Groups could be used in these custom reports, for traffic metrics, the funnel, ecommerce metrics, etc. However the powerful thing here is the range of dimensions to choose from. Common options would include:
- Device Category
- Channel
- Browser
- Country
- User Type
If you are capturing Visitor Type (prospect vs customer) in a custom dimension and/or Page Type in a Content Group (use Landing Page group to get Entry Points), this all gets more amazing.
Below is what you get as an output: a simple breakdown by stage of the funnel for whatever dimension/s you have selected. As a custom report, you would be creating it so you can drill down through dimensions to make it even more useful.
Knowing that your Conversion Rate is lower for segment X vs segment Y is not that valuable. Knowing that two dimension values behave exactly the same except for one stage in the funnel pinpoints where you have to take action.
Additional Points, Notes and Caveats
It must be noted that this technique works on the assumption that visitors must progress through each stage of the funnel. We know that this is not the case, especially when, for the above example, visitors could be entering the website with a persistent basket or creating a cart without viewing a product page. It is the job of the analyst to take these factors into account with any recommendations they make.
Further note that this is all session based analysis, as it is using Goals. For many businesses, visitors will take multiple sessions to convert. This approach is still useful though, in terms of seeing how far through the funnel the visitors proceed each time.
The overall technique is similar to an approach suggested by LunaMetrics back in June 2010. Their suggestion was to create a series of two step GA funnels for each stage of the website funnel and use Goal Abandonments for reporting. It would produce a similar report, although I prefer completion rates. It also means each stage needs to be based on pages, whereas this approach means you can use goals created from events.
Funnels do not have to be complex journeys. If you have a Contact Form on your website, it is more useful to know the number of form completions than to know the % sessions in which the form was submitted. This requires a goal for View Contact Form, a goal for Submit Contact Form, and a Calculated Metric for the completion rate.
Finally, we are looking into other Calculated Metrics as well. There is a list for content websites to calculate: Read Rate, Share Rate, Entry Rate, Engagement Score, etc. Watch this space for more ideas in the future.
