Over the years I have met many marketers that have Adobe Analytics implemented but are unhappy with their reporting. It’s such a shame to see so much time and effort invested in such a valuable resource that does generate return. A common mistake that is found over and over again is a poorly structured data model. Adobe offers a mass of customization that can easily be overwhelming. So let’s reel it in.
To get started, it makes sense to understand the difference between Custom Traffic and Custom Conversion variables. You can read about these here.
There are 2 types of variables that can be used to measure a user action; Data and Event (metric). Which means there are 2 different ways we can track a key engagement;
1. We can create Events for each action we want to measure and name them accordingly
OR 2. We can create key Event metrics and use a Data field to capture the event description
Because Data variables can capture unlimited values and an Event metric can only count 1 event, it makes sense to capture the value that can change; “text” in a Data variable and use custom Event metrics to measure key engagements, such as; View, Click, Download etc.
Another common mistake is to populate multiple Data variables with the same piece of data
Scattering data this way limits reporting capabilities. If we want to see all the activity for Login ID #123, we need to run 3 different reports!
This is why we always want our Data values stored in one place. Here is an example of a single report that uses key Events to expand the reporting capabilities of our data
Following these best practices for data modelling will ensure:
Less complexity
Consistent results
Data reliability
And most of all SCALABILITY (my word)
It’s hard to see the future and predict our analytics needs. Websites grow and content changes. If we build a reliable and scalable solution design, we don’t have to predict what may come. We’ll be ready for it.
For more tips on solution designing for Adobe Analytics visit this blog post
I welcome your comments and feedback
-Jane
Because Adobe Analytics is so highly customizable, we all want the benefit of experience to develop the best data strategies for our analytics reports. After creating analytics solution designs for well over 10 years, I’d love to share some of my best practices for robust and scalable analytics reporting.
Think about what your page name report will look like. How will you recognize pages and understand where they live on the website? By incorporating the page location into its name, you can easily identify your pages and create an intuitive view for all report users.
For Example:
Next, use custom variables to capture the new “folder” names as Site Sections for summary reports.
This level of aggregation can support;
1. Views of overall site traffic
2. Pathing analysis
3. Segmentation for custom calculated metrics
With the latest tally of 200 custom eVars and 1,000 custom events, we have a lot of customization to play with. Gone are the days of reserving a custom variable by checking the report suite and selecting next empty slot. With so much to work with, it makes sense to start with a plan.
Step 1. Plan your solution design by creating a list. Reserve “blocks” of variables for reports of the same category.
For Example: You may like to reserve the first 10 variables for key reports related to “Site Content”
Step 2: Use consistent naming and numbering for all associated variables.
For Example: If you are creating Link tracking in eVar8, it makes sense reserve event8 for the link tracking success metric (link clicks). So it’s best to use a template, like the one below, to draw up your plan.
It’s also a good plan to save a few extra variable spaces for report categories that can grow.
For Example: You may some day expand Campaign reporting by using Adobe Analytics Data Connectors to import metrics from your Ad Server. This may require reserving a custom variable for additional report data.
When we populate a single variable, we create the ability to report both, overall summary views and categorized reports. In this example, a single report will include all link activity on the same site. So, rather than creating a Header Link eVar and a Sidebar Link eVar, we place all the data in one Link eVar and include the link placement type in the data capture.
By compiling this data using a unique delimiter, we can use the Classification Rule Builder (in Adobe Analytics) to automatically parse data into classifications (reports).
Tip: 1. Scrub special characters from your data to avoid creating unexpected delimiters or sub delimiters. For example, you may compile your data using the pipe “|” delimiter as it’s least likely to be found in your data, but as it may still occur, you don’t want to capture it and create a new column in error. Also, special characters can mess with your classification rules and do not always convert correctly in your data. Example: Adobe reported this “Analytic’s:” when this was passed in an eVar/Prop “Analytic’s:“
2. Use a naming convention for the report names. The “Source Data” is what we pass to the variable, in this case eVar8. We use the full description for the source data “eVar#” and an abbreviation designates a classification “v#“.
Most of the time we will be reading classification reports as they offer nice summary views of activity.
+ The data structure supports data correlations for more detailed views.
The Source data is very helpful when running diagnostics or reconciling results. We just need to remember that Classification Rules require a few hours to process, so it’s good to keep this in mind when reporting current day results.
Adobe Analytics has been recognized, by Gartner Inc, as a digital marketing analytics leader for the last 3 years in a row. Bill Ingram, vice president of Adobe Analytics Cloud, claims close to two-thirds of the Fortune 100 leverage Adobe Analytics for customer intelligence. Adobe offer a robust reporting with in-depth pathing and segmentation but not all marketers know how to use it to its full potential.
The key to understanding Adobe Analytics starts with understanding the differences between Traffic and Conversion reporting and why we need both. Adobe divide analytics into these 2 key areas as, Traffic reports offer insights into visitor behavior, such as pathing, time on site and site content visited. A Conversion report will advise of specific actions or success events that occur during the visit.
Conversion is measured in both; predefined variables (such as Campaign and Product) and with Custom Conversion Variables (known as “eVars”). The term “eVar” is a legacy name and is short for “eCommerce Variable”. Back in the early days of analytics, it was thought that eCommerce events were the only significant events to track. Today we track File Downloads, Video Views, Internal Searches and many other key website actions in eVars.
So how do we use custom conversion reporting? We pass the event description to an eVar and count how many times the event occurred using a custom Success Event. Usually these are paired together.
But we don’t stop there. A robust Adobe Analytics implementation will have both Primary and Supporting eVars.
A Primary eVar is one that has a related event. File Download (eVar10) is primary as we also have File Download (event10)
A Supporting eVar is used to report related information at the time of the event. Example: Site Section (eVar1). By passing the Site Section name at the same time a File Download is captured, we create a new data relationship that allows us to report total File Download events by Site Section.
We can also go a step further and break out Site Section by File Name to see which files are downloaded in each Site Section. Because these values are passed to Adobe Analytics together, they can be reported together.
So now you might ask, why am I also passing the File Download value into a traffic variable? Ex: File Download (Prop10). Custom Traffic variables, “Props” are traffic “property” variables.
We use these in combination with eVars to measure overall engagement.
For example, the conversion report metric (success event) reports the total raw count of each defined action. Such as the total downloads for each file. This may tell us which files are most popular but doesn’t account for repeated downloads in the same visit or by the same visitor. As Visits and Unique Visitors are traffic metrics, we want a “traffic” report as well. You may note that Adobe allows us to add traffic metrics to conversion reports, but to get the most reliable results, we always recommend reporting traffic metrics on traffic reports, such a custom traffic variables.
So now you can see Adobe Analytics structures data. What are some of the custom events/engagements you want to track?