A few days ago, Google Analytics announced it was introducing ‘calculated metrics’. When we got chatting about this in a planning meeting recently, half of us said, “Yes! That’s cool”, the other half weren’t so enamoured.

I was in the excited half of the group. And I immediately wanted to rush into GA, setup my 5 calculated (or compound) metrics and sit back and wait to be amazed.

But it made me think, is this a good thing or could it pose us marketers a serious headache?

Let’s start by looking at calculated metrics and why they could be a good thing.

What Are Calculated Metrics

Calculated Metrics (commonly referred to elsewhere as Compound Metrics) are the product of combining two existing metrics.

The issue we’ve had with Google Analytics is that whenever we needed to combine two metrics, we had to export raw data into Excel and do the maths there.

Now, with this new release, Google Analytics is able to calculate the product of two metrics. Two examples, given on the Google Help Doc are:

  • Revenue Per User. To calculate the average revenue generated by each of your website users, you would now have a calculated field that worked out the sum {{revenue}} / {{users}}
  • Currency conversion. This is handy for businesses that operate in a number of regions. For each country’s reporting view, you can use a calculated metrics that converted the global revenue figure into the local currencies. For example to convert GBP to EUR would look something like {{revenue}} * 1.27.

Why Calculated Metrics Are Cool

Basically, it’s a time saver. Being able to calculate metrics directly in Google Analytics, saves us marketing types the time to export to Excel or to write complicated API calls.

I also think it’s a good move for Google Analytics. Compound metrics have been available in the analytics products of their competitors for a while now. And I think Google have seen Adobe Analytics start to become the default option for enterprises and they are motivated to change this. If this competition results in a better Google Analytics offering for SMEs, then that’s a good thing.

Where could it go wrong?

Marketers are an inventive bunch. We’re creative and that might be our analytical undoing.

It’s now possible to measure almost anything in Google Analytics. The Measurement Protocol and Custom Dimensions and Metrics widened Google Analytics’ scope to include almost anything.

Focusing on web behaviours, some of the useful things we’ve created custom tracking for are scroll depth, form completes, social shares or comments.

The danger is now that marketing can combine these metrics, we will.

For example, we could create a calculated metric that multiplied {{number of comments}} * {{scroll depth}} and produced a new metric called “Reader Engagement Index”.

What on earth does “reader engagement index” mean? How do we influence it? As the component parts are independent, one can rise and the other can fall and our ‘reader engagement index’ would be the same. In other words, by combining two independently useful metrics, we created a useless one.

So, the word of advice is that calculated metrics are great, helpful and can save you time. But they should be treated with caution.

How to avoid creating dud metrics

Understand your users

First and foremost, understand your users. What do you want them to do? What actions do they take on the way there?

Your metrics should measure those interactions.

If you need your users to complete a five field form to complete your purchase, a handy calculated metric would be ‘Form Fill Percentage”, where you calculate the number of fields completed as a percentage of the fields available.

Understand your business

What does your business need to know to manage performance? The examples given above around revenue per user, or, currency conversion would be very useful.

But what else does your business need to see and currently calculate manually?

Create Metrics that are actionable

I believe the most important question to ask yourself when creating a metric is “what am I going to do with this insight?”

If you’re a publisher and ad revenue is crucial, you might be very keen to ensure that your users scroll right to the bottom of your content so you can load more ads.

Using a simple scroll-depth metric, you’d be able to make editorial and design changes to get more people reading further.

This principle of “what am I going to do with this insight?” needs to carry over into your calculated metrics. Before you create them, make sure you know what you’re going to do with the information.

If you’d like help with figuring out what to measure and how to do it, get us on @rocketmill and we’ll be happy to help.