Measurement planning for useful digital analytics
I’m a digital strategist and so I spend a fair chunk of my working week with my nose in Google Analytics, plus a few other products that allow us to measure patterns of use and behaviour in digital products and services. There are so many brilliant tools around for this. But how do we make measurement actually useful?
This blog post talks about how we approach digital analytics for ourselves and our clients, it looks at a few tools and some useful mindsets. What it doesn’t do is talk about how to optimise and improve based on your analytics - we’ll cover that in a later post. I’m using ‘website’ in this post as a shorthand for really any digital platform, service, product, application and so on.
Why measure anything anyway?
Let’s start with the best question - why even do this at all? Let’s set digital analytics to one side for a moment, you’ll find that whatever your business or organisation, a bunch of stuff is being measured already. Your sales manager is measuring whether they have enough new business enquiries from your website to hit their sales targets. Your events team are measuring whether or not they get enough sign-ups to their events. They might not be using anything special to do this, maybe just a spreadsheet, maybe even just anecdotally, but they will be measuring.
An organised digital analytics programme simply aims to formalise this, and then build on it.
You’ll notice that neither of the examples I’ve given seem to have much to do with the traditional ‘web’ measures of Pageviews and Bounce Rates and the like. One of the problems with digital analytics is that these kinds of metrics - whilst extremely useful in context - easily draw attention away from our real aim, which is to see how well the website is doing the things that it’s meant to do for your business or organisation. Websites are supposed to do stuff for you, yes?
This is where measurement planning comes in, and measurement planning is our primary tool for making measurement useful. Here’s how we put a measurement plan together.
1. Tie your measurement to business goals
Before we do anything else we identify what is useful to measure. This always begins with the organisation or business’ objectives/activities. To show how we document this, let’s use a fictional client: a medical royal college:
First we list out their core organisation objectives, in this case there are 3. You may have more or less.
Then, we list out the practical things that the website does to support those goals.
Start simply and broadly. Think about your most important organisational goals. Think about how the website is meant to directly support them.
You can always add more later on.
2 - Break it down, keeping it human
Now let’s take each one of those ‘things the website does to support the organisation’ and get into the detail of what we want to measure.
Back to our fictional client again. One of its objectives is to provide CPD training in its area of specialism to 10,000 doctors each year. The website supports that by listing the CPD events and allowing online booking. To know if it is doing a good job of this or not we probably want to know stuff like:
- Is the events section signposted clearly and prominently enough, are people finding it?
- Is the summary information about the CPD events clear, useful and enticing, so people want to read the detail and attend?
- Are the booking calls-to-action clear, and is the booking process easy? Or are people dropping out?
- Do people who book through the website then go on to attend the training?
We find it most meaningful to ask these human questions first, not “what is the bounce rate on page xxx”. It helps us know if we are measuring useful stuff. THEN we can identify the specific metrics. In our example it makes sense to simply measure:
- How many people are viewing the events section
- How many people are viewing individual events
- How many enter the booking process
- How many complete the booking process
- How many attend the event having booking online
Here’s how that looks in our plan:
Quick word about custom dimensions
Dimensions in analytics are a way to segment and interrogate your data - dimensions are properties of users, pages or sessions. For example, the acquisition source of a session is a dimension. You can see in our example plan we have a column for ‘custom dimensions’, because we can add our own in Google Analytics and this can be really useful. Let’s say in our example, a user could be either logged in or not logged in to make their booking and we want to know if users behave differently based on that. And, the CPD events fall into a number of different categories and we want to see if some categories are more successful than others. If we pass this dimension data into Google Analytics along with the pageview or the custom event, we can start to understand these patterns.
3 - Smart implementation
There are loads of great software tools out there to help us measure things. For our example here, Google Analytics will do most of the work for us:
- How many people are viewing the events section (pageviews of the event section)
- How many people are viewing individual events (pageviews of each event)
- How many enter the booking process (a custom event)
- How many complete the booking process (a custom event)
I’m a fan of custom events being captured in Google Analytics because they are versatile, and easy to interrogate, and Google Tag Manager makes them ridiculously easy to configure, test and adapt. Tag Manager is now our default tool for implementing digital analytics tracking.
In our example can’t measure event attendance through GA though, so our last item would be:
- How many attend the event having booking online (event register on the day)*
That’s right - don’t forget that our actual organisation goal here is to get people to attend the training. So we want to know if they turned up on the day!
Here’s how we show the implementation method in our plan.
Another favourite tool is Hotjar which we find complements Google Analytics very well. Hotjar is great for ‘on page’ analytics, helping reveal what users are actually doing when they arrive on pages of your site. We use its form tracking capability a lot, which tells us how users behave with each field of a form and where they may be dropping out.
Let’s say in our example we find that plenty of people enter the booking process but relatively few complete it. We can use Hotjar to help find out which parts of the booking form are causing problems for users and making them abandon their booking. Using the right tools for different parts of your digital analytics programme - and not shoehorning everything into Google Analytics - is important.
4 - Reporting and review
Our measurement plan sets out expectations about how we’re going to report on and review our metrics. One thing you’ll know if you’ve spent any time inside Google Analytics is that you can use up a lot of time rummaging for the numbers you need, so it helps save time later on if you think ahead to what specific reports will help you.
Here’s how we show that in our plan.
I tend to set up custom Google Analytics reports to pull out the key numbers and do some basic interrogation, and then use the default Google Analytics reports if I want to drill down to get some deep insight. Our plans don’t go into the detail of which dimensions and metrics we’ll show on each report, but you can if you want and this would make sense if the reports are potentially complex. On the whole though I find custom reports so quick and easy to set up and configure that these just get updated on the fly if we find we need new information.
Dashboards are an under-used feature in Google Analytics but can make your reporting very effective and efficient. A dashboard is simply a grouping of reports. I like to create a dashboard for each of the business goals that were identified at the beginning of our measurement planning. Why?
Let’s look again at our example around CPD events. To get a complete picture of what’s going on we really need to look at our different metrics side-by-side to get a sense of how our users are moving through this journey: where are the drop-offs? Dashboards give you that capability - you can then apply segmentation or filters to your whole dashboard.
If you want to get REALLY down and dirty with some custom reporting you can’t get much better than Google Data Studio which is a free tool with native Google Analytics integration. Just like with our Google Analytics dashboards, we can set up side-by-side metrics reporting but more powerfully still, we can create reports with custom calculations to automatically show us things like % drop-off rates between steps in a journey.
Schedule your reviews
If you can, also use your measurement plan to set out your schedule and teams for reviewing the numbers. It doesn’t matter if this needs to evolve in time, but successful digital analytics programmes tend to work best with a bit of rhythm to them (review, plan, do), while being ad-hoc and reactive can mean you’re only addressing the needs of the stakeholders with the loudest voices.
Summary - and something to watch out for
Using a measurement plan is a quick and simple way to make sure you’re being useful with your digital analytics. And they really are quick if you know what your organisation or business’s objectives are: most normally take only a couple of hours to put together.
One thing worth remembering is that once you have analytics tools installed on your site they’ll be capturing all kind of behaviours which aren’t covered by the scope of your measurement plan. Use your measurement plan to give your analytics programmes direction and purpose, but don’t forget to check in from time-to-time with your more general site stats as well.
Here’s a link to our fictional client measurement plan, feel free to take a copy to use for your own site.
If you’d like to know more about our digital strategy offering which includes work on digital analytics programmes, get in touch!