Three foundation ingredients for Google Analytics attribution
From a marketer’s perspective, attribution is one of the most powerful tools available from web analytics platforms such as Google Analytics (GA). Google defines an attribution model “…the rule, or set of rules, that determines how credit for sales and conversions is assigned to touchpoints in conversion paths.”
By analysing attribution data, we can find insights about return on investment and the relative performance of our marketing activity. We can then use these insights to inform future channel strategy and budget allocation.
The standard Acquisition reports in GA offer highly valuable insights into marketing performance without requiring advanced user skills. For many brands this level of insight is enough for your long-term purposes. For others it is the start point only – for you Google Analytics offers several other more advanced methods of modelling attribution, which will be covered in a future post.
Regardless of the sophistication level of your attribution analysis, you always need to get the foundations in place first.
This post focuses on the foundations of channel attribution in Google Analytics, ensuring you are set up to collect accurate and usable attribution data for use with Google Analytics acquisition reports.
Google Analytics acquisition and attribution reports
To provide some context on this topic, let’s start with the end data that we use.
The Acquisition reports in Google Analytics are the primary out-of-the-box reports for understanding how well particular marketing activity is driving visits, engagement, and conversions on our websites.
Here we can review attribution data at different levels of detail, starting at high level Channel Groups, but also drilling through to the granularity of Campaigns and Keywords.
The standard reports also enable us to see these campaign dimensions lined up against Acquisition, Behaviour and Conversion metrics, so we can quickly assess the relative performance and return on investment of our activity.
Before making decisions based on this data it’s important to understand how it is generated and ensure it is accurate.
The three key ingredients to Google Analytics Attribution
There are three interrelated ingredients required for successful attribution reporting:
- Campaign tags (also known as utm tags or utm code)
- Attribution dimensions: (Channel Groups, Sources, Mediums and other detail)
- Reporting metrics
1) Campaign tags
With the correct tag set up Google Analytics can capture information that is attached to each link click bringing traffic to the site. This information can include where the link was clicked from as well as which campaign, specific content, keyword or other information of your choice.
There are several methods that GA uses to collect and collate this information:
1.Referral traffic . For any traffic coming from another site Google Analytics will record that referring property. For example, for an untagged visit generated from facebook the source/medium will be facebook/referral. While this enables you to automatically report the source and medium information, you will not get any further detail on campaign, content or term, so sometimes applying a tag with more detail is preferable.
2. Integrations and automated campaign tagging . A direct integration of e.g. your Adwords account or Search Console with Google Analytics, will automatically provide the relevant campaign parameters to GA. Integrations can be set up in your GA administration interface.
Many email and automation tools now also offer a tick box option to automatically generate utm tags that Google Analytics will read, removing the need to manually create these.
If these integrations or automated tagging options are available to you, then I recommend you use them.
3. UTM code. A utm code (sometimes referred to as a utm tag or a campaign tag) is a snippet of text that is appended to your campaign landing page URL and carries information that Google Analytics can recognise and organise on the link Source, Medium, Campaign and more.
A utm tag is typically made up of between three to five parameters: Source, Medium, Campaign, Content and Term.
Creating utm tags is made easy with the availability of some excellent ‘paint by numbers’ tools.
My post ‘UTM tags in a Nutshell’ covers everything you need to know.
Any site visit that does not have a recognisable Source or Medium via one of the above three methods will be attributed as a ‘direct’ visit. For example an untagged link click in an eDM will be attributed to the direct channel in Google Analytics with no information available on website engagement or conversion driven by that eDM link.
2) Attribution dimensions: Channel groups, Sources, Mediums and more
In Google Analytics metrics are ‘counters’ and dimensions are ‘descriptors’. The three key parameters that you collect in the utm code or integration (Source, Medium and Campaign) are reflected in the main dimensions in GA attribution reporting.
This is why it is so important that we name and collect this information consistently and accurately, so that credit is allocated accurately.
Our data and analysis is also made more manageable through the ‘rolled up’ Channel Group .dimension.
A Channel Group gives us the highest level of ‘at a glance’ information on the relative performance of our marketing activity. If I’m looking to analyse and optimise tactical activity I would work at a more granular level of source/medium or campaign, however Channel Groups are excellent for concise management reporting.
A Channel Group is not one of the parameters in campaign tags, but rather an even higher level collection of similar marketing activity usually collated using a combination of Mediums and /or Sources. For example the Paid Search channel may bring together any activity with mediums of ‘cpc’ or ‘ppc’ or ‘paidsearch’.
Google Analytics provides an initial set of Default Channel groups out-of-the-box so you don’t need to manually set most of the rules and information for source, medium or channel groups if you don’t want to.
The Default Channel groups available out-of-the-box in Google Analytics are:
- Organic Search
- Paid Search
- Other Advertising
- Direct (this is activity with no known source or medium)
In order to accurately direct your utm tagged activity into these channel groups, you need to either follow the exactly naming conventions laid out in the GA out-of-the-box rules for each Channel group.
See Google Analytics documentation for up to date out-of-the box rules for Google Default channel groups.
You can also tweak the Channel Group rules to suit your preferred naming conventions, internal language and reporting needs. For more on this see my post on editing Channel Groups.
While most commentary focuses on using Attribution data for insights into bottom of funnel conversions such as sales, I also like to analyse other top and middle of funnel metrics. This ensures I understand the role of each marketing activity across the full user journey, from acquisition (site visit) to behaviour (engagement with the site) and conversion (goal).
User acquisition metrics: Attribution of marketing activity against metrics such as sessions, unique users and new users is helpful. At an absolute level, which activity is sending the most visitors? Are there patterns as to when this happens or which devices they are using? Am I attracting new site visitors?
Engagement metrics: Once visitors get to my site, what are they doing? Is any activity driving a disproportionate number of bounced visits? Which activity drives high quality visits that stay on my site, and visit pages and sections that I want them to get to? Are they performing micro goals that indicate they are a visitor with potential, even if they haven’t completed a major conversion yet?
Conversion metrics. While attribution can be used for the site visits metric so you know how people got to your site, it is also useful to know which channels are driving conversions. In this case, you will want to have goals set up, so you can relate the marketing activity to the successful outcome. You may find that some channels drive a lot of site visits, but ultimately don’t deliver the results you need, while other activity may send a lower quantity but far higher quality of site visitors from a conversion perspective.
Even if your site does not have a checkout function, you may have lead generation actions such as ‘register your interest’ or ‘sign up for newsletter’ that you count as a conversion. Goals are not automatically set up in Google Analytics. See the Google support documentation for information on setting up Goals in GA.
Accessing attribution data and going to the next level
Beyond the standard acquisition reports that you can access in the Google Analytics reports interface, you can also create any combination of attribution related dimensions and metrics in either a Google Analytics custom report, or in Data Studio, enabling you to regularly access the exact mix of data you need in the format you need it.
Once you have mastered your tagging protocols and standard Attribution reporting, there is much more that you can do with attribution in Google Analytics, such as multi-channel attribution, understanding assisted conversions, attribution modelling or working with the beta ‘Attribution’ interface.
When we think about the insights and decisions that are made based on this attribution data:
- understanding what is driving success,
- assessing return on investment,
- future channel strategy and budget allocation,
this highlights the importance of ensuring the data ‘in’ is accurate.
Good luck getting started with attribution yourself, or feel free to get in touch for assistance.
Annie Clare Consulting helps brands take the guess-work out of decisions, through data-driven insights, strategies and optimisation.
Get started with free Google Analytics audit and plan, to help step up your data-driven decision making and meet your customers’ needs.