Basic Tips for Using Adobe Analytics Campaign Tracking Codes
By Rob Saunders, Practice Lead Solution Architecture and Todd Deno, Principal Consultant Analytics
One of the fundamental reasons that digital marketers rely on Adobe Analytics is get actionable data that will allow them to make better decisions. Knowing a robust amount of information about your campaign level activities enables a conversation around, what marketing activities are most effective? You can then make strategic decisions around where and how to invest your marketing dollars.
When getting started with the process of assigning tracking codes to marketing campaigns, the most important step is to build in logic and structure for how the tracking codes will be defined. It is technically possible to use a randomly assigned alpha numeric value as a tracking code, but the process of analyzing and maintaining this becomes time consuming and cumbersome. The main reason most companies go with a random value with no structure is that this is the default method of their advertising agency to assign tracking. While this makes it easy on the agency, it makes it difficult on you. Remember, you are paying them, THEY work for YOU, so make your life easier.
Building logic and structure into your tracking codes will require more upfront planning, but it will make your life much easier going forward. So what do I mean by logic and structure? It’s simple: your tracking codes should follow a set naming convention to help you identify what the code is and follow a structure that is consistent across marketing channels. For example, say you are running a paid search marketing campaign in Google, if you assign a tracking code similar to “sem-google-brand-broadmatch-123456” you would be able to identify what this tracking code is simply by looking at it. But, the benefit of this strategy does not end there. Since we have a pre-defined name for paid search campaigns (i.e. ‘sem-‘) we can setup marketing channel processing rules in Adobe Analytics to automatically assign all tracking codes starting with this identifier as ‘Paid Search’. This will roll up all of your paid search efforts for you to analyze at an aggregate level on both a first and last touch allocation.
But wait, there’s more. Beyond marketing channels, you can also construct classification rules that will auto classify your tracking codes for you. That’s right, no need maintain and upload SAINT* files! Similar to the logic for marketing channels, we can classify all paid search efforts as ‘Paid Search’, but we can take this even further and use regular expression to pull out various sections of the tracking code (this is where the structure comes into play). So, with the tracking code “sem-google-brand-broadmatch-123456” we can use regular expression to break this code into five parts (sem, google, brand, broadmatch, 123456). So now we can see all of our Google, brand, and broad match keyword campaigns at an aggregate level.
And, we aren’t done yet. Adobe Analytics also allows you to create sub-classification (classifications of classification). This allows you to use both the classification rules and SAINT together to get even more out of your tracking codes. Take the section of the tracking code example “123456”. This could represent campaign creative which would be something we could never setup a rule to identify. But, if we use rules to classify this under ‘Creative’ you can build a sub-classification(s) of this to be “Text”, “Ad Size”, “Placement” etc… and use SAINT to upload the actual values associated with this number. So, now you can see that your campaign that had a Call to Action, “Download Report” in a “300×250” banner on the “Home Page” drove a 30% click through rate.
This idea of sub-classification can also work for other sections of your tracking code. Take the example listed earlier, google, brand, and broadmatch may make sense for paid search but doesn’t necessarily make sense for email. You could simplify your tracking code to contain [channel]-[campaign code]-[creative code] and use sub-classification to break this down further, but the more logic you build into your codes the less reliant you are on SAINT files, which equals more time analyzing and improving and less time maintaining. I am certain that spending more time finding insights and less time on reporting was a resolution everyone made for 2015.
*SAINT stands for SiteCatalyst Attribute Importing and Naming Tool
