5 Steps to Actionable Metrics | Converge Consulting

5 Steps to Actionable Metrics

1. Build a Measurement Plan

This may seem like a no-brainer to some, but this step is the one that is most often skipped and causes the biggest headache. So take some time to map out what exactly you want to track on your website and also be sure to ask yourself “Why do I want to track this? What actionable insights can I gain from this metric/dimension?”. This is important because even if you have the ability to track every click on your website (and with GTM you certainly can), doesn’t mean that you should. Think about what are the most important metrics that you can take action on and make a positive change for your department. You can tell your boss “50% of users clicked on hamburger menu”, but what does that actually say about your website? Once you have all your KPIs listed with important information like your overarching goals and what metrics you want to get from them, it’s time to map out your tracking into any analytics platform (like Google Analytics).

2. Audit Your Current Analytics Set Up

Now that you have a Measurement Plan in place, you’ll need to look into how your current Google Analytics (GA) and Google Tag Manager (GTM) are set up. Although it is a very common misconception, GA and GTM are not the same thing. GTM is a tool that allows you to create, edit, and manage JavaScript tags from Google and third-party tools on your site. While GA is an analytics platform that is focused on collecting and reporting data about your site. This is important to know for this next phase as you will need to dive deep into your current implementation to see what and how you are tracking.

  • NOTE: If you are still only using GA, then this is the time for me to tell you that you are seriously missing out on a very powerful tool for marketers to control their tracking without changing the actual page source or knowing very much HTML/JavaScript by using GTM! Bonus: GTM is a more effective way to handle JavaScript tags since they only fire tags when they are configured to trigger. This means that you don’t have to implement certain tags on only a few pages or removing tags on certain pages since all of that configuration can be done through GTM. There are definitely still instances where having a developer’s help is needed, but it’s a pretty user friendly interface for basic tag implementation.

To start an Analytics Audit, I dig into GA’s settings and export large reports from GA in order to look at all the pages, events, and campaigns for at least a year. From there I have a series of checklist of questions, but here’s a brief summary of the reports and thinking for each of the 3 main reports.

  • All Pages Report – This is a great tool to see popular pages but also to look at the overall folder structure of your site to see if makes sense and is consistent. If there are pages that no one visits, then you can start asking questions like “Is this page still relevant?” or “Should we even still have this page up?”. If there is a good reason to keep the page up then, “Are there opportunities to link to this article to help with its visibility?” or “Should we rework the menu selections?”
  • Top Events Report – I use this report to see if what a client finds important or if they find everything important. Just because you can track everything doesn’t mean you should and you’ll see why I say this if you look at your Top Events Report and you have 50+ Event Categories. For larger Universities that are sharing a GA Property, this isn’t uncommon since each university tracks their events differently, but I always try and make it as simple as possible. I like to imagine that a brand new employee is sitting down and has to try and understand these reports. Would they be able to find the number of form submits in a few clicks?
    • NOTE: I tend to shy away from using URLs in my Events since GA automatically tracks which page an event occurred on and most users don’t really read through full URLs (the exception being outbound links). However, if this method is something that has been in use for awhile, then that is something to consider since historical data can’t be changed in GA.
  • Campaigns Report- One of the most useful dimensions in GA since it’s the easiest user defined variable and the best way to really gain understanding of your marketing efforts. However, since it’s a user defined variable it’s only useful if the naming is descriptive and consistent. GA is case and space sensitive so if there are any differences in your naming then they won’t aggregate and it won’t be super easy to understand which campaign was most effective. That’s why we tend to recommend that client’s use a UTM Builder, keep naming as consistent as possible, and keep a record. This is especially helpful if there are multiple departments/colleges sharing a GA property.

All of the questions and information from the Analytics Audit will really help when you start mapping events from the new Measurement Plan into your Implementation Plan. Which leads me too…

3. Create an Implementation Plan

Now the true benefit of the Measurement Plan and Analytics Audit will shine as you are able ask more detailed questions as you figure out the most effective way to track the new KPIs into GA (or any other analytics platform). This allows you to really iron out any kinks with naming conventions for all the components (like pages, events, campaigns). Please note that any changes you make in GA will not change historical data so changes in namings might mean you will have to manually compare old data to new data. My opinion tends to be that that’s a sacrifice I make to be able to really get actionable and readable data, but if that’s not the feelings of other departments or the university then you will have to discuss or set up your own instance of GA and/or GTM. The Implementation Plan will help identify how the KPIs need to be set up in GTM tags and how to set up properties and/or views in GA. Once your plan is in place you can start acting on it for GTM tags, but for the GA set up I have a few notes in the next steps.

4. Set Up Google Analytics Properties, Views and Filters

Setting up a GA is not as simple as adding GTM code to your site since it’s never really a one-size-fits-all answer. There are pros and cons of using properties and views, especially for a setup like a University where you are often grandfathered into whatever approach was put in place years ago. On a very simple level, a GA property uses the same tracking ID; therefore, if you have multiple tracking IDs for GA then you have that many properties. In my experience, most clients have at least one property that is for the entire university and then multiple others for each department/college.

For each property the best practice is to have at least 3 views with the following suggested filters:

  • All Website Data View – Unfiltered – Make sure that you have one version of the data with no editing or filtering. This is especially helpful with troubleshooting and debugging.
  • Filtered View – External Traffic Only – With this view you should try and filter out any internal/consultant traffic usually by IP, but if you share IPs with your students and consider them a key audience for your site then you may want to keep them or separate them out into a new view for just Internal Traffic.
  • Test View – Internal Traffic Only – This view should be filtered to only include data from your team/consultants so that they can see their traffic for testing purposes and have it separated from the filtered view. You can also use this view to test new filters, goals, or other GA settings to make sure you don’t break your current set up.


5. Get to Analyzing and Finding Insights!

You have created and implemented your Measurement/Implementation Plan, so you can now move onto the most important part about tracking — making actionable decisions based off of insights in the data! Be sure to remember that sometimes that decision will be to update or change the way the tracking works and for that I would suggest starting at the step 1 and documenting and discussing each step and change. Going through this process will help empower your team/department into better understanding your data and by proxy, your users!

Maddie Cantrell
Maddie Cantrell
September 13, 2018