GA 101 - What are Views

Introduction

Journey into Google Analytics (GA) for the first time and there are so many different options it can be overwhelming. GA will have reports on traffic, revenue (if you’re an ecommerce shop), user behavior, top events and so on. The last thing on your mind is whether or not you’re in the right view. In fact, most people new to GA don’t know about different views and assume they are looking at the correct one. As you’ll see in this article, being in the right view can significantly impact the information and data you see. We’ll discuss what views are and which views you should have on all your GA accounts. 

Google Analytics - Admin

If you are not familiar with Google Analytics or Web analytics, check out this post to familiarize yourself.

Figure 1 - Admin view of Account, Property, and View in GA

GA is managed in a hierarchical role Account -> Property -> View. The account will typically represent the brand and is overseen by the GA account manager. It is the highest form of privilege within GA and all access granted at the account level will trickle down to the property and view. Typically, for single brands there will be only one account; however, a large brand or GA Admin overseeing multiple accounts will have more than one account listed. 

Next, the property is where you determine where to send your data with GA. It contains the GA Tracking ID along with information on the number of hits for the account and whether to attach Google Search Console to your GA instance (yes this is highly recommended). It’s also where the code snippet is to load GA onto your site; however, I strongly recommend utilizing Google Tag Manager. 

There are a lot more items to cover in the property, but the final piece I will touch on is that with properties you can create multiple instances to track different sites or different parts of the site. At Stance, I utilized this to track our recurring subscription business as the business model differs greatly from the behavior of regular orders. Another instance of when to use multiple properties is for a staging/dev property so all data in those test environments do not flow into the regular GA property, such as test orders etc. Otherwise those test orders would show up as part of the regular revenue. There are a lot of creative ways to utilize the property within GA, just know that unless you are on a GA 360 account, a lot of the settings do not carry over from property to property, particularly things such as custom dimensions and metrics which will need to manually be transferred in order to be tracked.

Views

The best way to describe views is how we want to see the data. Within views there is the ability to apply filters, set goals, determine if you want to enable enhanced ecommerce (recommended for ecommerce sites), alter the channel grouping or set custom alerts. Views allow us to get creative with the data to see what we want to see without impacting all the data in the account (more on that later). Therefore, if you wanted to create a view just for FB traffic, you could create one without having to track with a different ID as you would with a property. The main difference between a view and property is that all data from the view comes from the parent property. Thus if you have different properties, the view will only ever be able to get data from the corresponding property. 

One important tidbit to know about views is the fact they are not retroactive! Therefore, data will only start collecting from the moment the view is live. This also applies to filters or query parameters applied to the view. Therefore, it is always good to have a goal in mind when creating a view so it is serving the right purpose. 

Views to Have on All Properties

One of the first items I always check when working with clients on their GA Accounts is to make sure there are at least these 3 views: Raw (or often labeled unfiltered), Master (primary view account), and Test views. These 3 accounts are critical to data integrity within GA and help serve as a checks and balances in GA. To see what Views are available on your property, click on the drop down at the top right left corner next to the Analytics name. This should open up a drop down menu with your account, property and the list of views available within the property (shown in fig 2).

Figure 2 - Accessing different views

The Raw View, oftentimes labeled as Unfiltered, serves a great purpose of essentially just saying “Give me all the data that is being tracked on my site into GA”. This view is created by default when creating a GA account for the first time. What I’ve found useful for the Raw view is to find any potential spam/bots from the data, along with a validation for the Master view. The Raw View should always have a greater count than the Master View, otherwise, this is an indicator something is not right. The most common mistake I’ve seen is having the Raw View as the default view in the Property Settings (fig 3 shows where to find this in the property). This means that by default, users will land on the default view, where they could be making decisions utilizing data from bots and spam. Yikes! By default, this should always be set to the Master View.

  Figure 3 - How to check your Default View in GA

Test View

Before jumping into the Master View, in my personal opinion the Test View should always exist for every property. Remember how I mentioned earlier that View data is not retroactive? Well, the same rules apply if a bad filter or improper configuration is put in place, leading to potential loss of data within the view that can no longer be retrieved. Having a Test view to make changes and updates, observe and validate there are no issues is a surefire way to protect your data as there is no way to recover data. 

An example I saw was an improperly configured filter to only include traffic from a specified hostname (very common filter to filter out spam). However, as the regex was written improperly, the traffic to the site dropped off completely without any notice. The early assumption was that maybe a developer wrote some bad code and the tag was no longer firing or something was updated in GTM and it was impacting tracking (all common and fair hypothesis). After a few days of investigating that showed little progress, we thought to check GA to see if anything was updated (typically forgotten as not many people would make changes here) and of course it was the improperly configured hostname filter. It was promptly fixed, but the data during the timeframe where the filter was active were gone forever.

While that is a worst case scenario, it could’ve been mitigated by having a test view. Because as the name implies, if that had happened within the Test View, it is alright since it is not the main view people are making business decisions on. However, to have it impact the Main View, definitely will have a trickle effect on Ad Spend and campaign performance directly tied to revenue. If you’re still on the fence about whether or not to include a test view, do it and save yourself the headache of having to explain to an executive why there are days with no data or missing data.

Master View

When thinking about the master view, think of a salmon, the whole fish would be the “Raw” view as you have all the parts: head, tail, organs, fins, etc. However, most people don’t consume the entirety of the fish all at once. By slicing and dicing the fish, parts can be used for various recipes. The master view represents only the parts of the data (or fish) which we want to consume, thus removing all the unnecessary parts.

What goes into the master view that separates it from the Raw view? The first couple of items may sound like no brainers but it’s a lot more common than we realize. First, bots and spam hitting sites are not normal traffic. Their behaviors are meant to crawl the entirety of your site, mining for information. Basing decisions off this type of behavior would easily lead us astray and some of these bots (i.e. Google), we ask to crawl our site for SEO/SEM. Spam is another piece to exclude as these are not real users. This can be validated by looking at the hostname, which should always be your domain and not others like those shown in fig 4 below. Having a lot of spam representing “users” on the site can greatly skew bounce rate, sessions, and avg. session duration.

 Figure 4 - Examples of Spam to Website

Another underrated piece of the Master view is grouping content (url’s) to make it easier to parse the data. When looking at some of the top landing pages, most often you’ll see the top 10 links and leave it at that. However, due to query parameters, links can be broken up into many small parts which as a sum may boil up another link. When fixing this for sites, I’ve seen some links which were outside the top 20 suddenly appear as a top 10 just by cleaning up the query parameters. Having this type of information can greatly impact where our priorities lie in terms of landing page optimization and user experience design. 

For those not aware of query parameters, they’re typically found in the url following a ?. Here’s an example of a query parameter and how GA would track it by default. Page Path /mens-best-sellers/?prefn1=refinementColor&prefv1=BLACK&prefn2=sizefilter&prefv2=M&prefn3=underwearfit&prefv3=Boxer%20Brief.

To most users, this would be difficult to parse. To those more familiar you could figure out the page is filtered by color (black), size (medium), and underwear type (boxer brief). The query parameters in this case would be prefn1, prefv1, prefn2, prefv3, prefn3, prefv3. Now, say we decide to exclude those query parameters from tracking, more content would now be grouped under /mens-best-sellers/. This change could improve the numbers of /mens-best-sellers/ to suddenly be a top 2 or 3 page instead of a top 10, which would really impact the way brands campaign or the site navigates people to content. If you’re concerned about still wanting to track some of these query parameters (in this case filters), I’d highly recommend utilizing event tracking to keep the data easier to digest. 

Conclusion

Understanding the different Views in GA is no easy task. Unless you get training or spend time separately to go and learn Google Analytics, the different views won’t even cross our minds. Check whether there are 3 views set up (following the suggestion in this article) and make sure to set the Master View as the default. Just as you wouldn’t eat a whole salmon, by cleaning up the fish, we can consume it in the manner which we please, whether it be as a steak or as part of a sushi dish. Views allow us the ability to look at this data in any manner we see fit for our data-driven quest.