Loading

Analytics & Target Best Practices for Analysis

Analytics & Target go together like bacon & eggs. It just makes sense!

If you've heard these comments around the water cooler, your organization could benefit from leveraging Analytics as a data source for your Target activities (A4T).

What's in it for me? Why use Analytics?
  • While the same data is available in both interfaces, the Analytics UIs are best suited for deep analysis, advanced segmentation, and data visualization
  • Analytics metrics & segments are “always on” so there's no advance planning or setup required which saves time & helps gain efficiency
  • Analysis Workspace has sensei capabilities such as Anomaly Detection, Contribution Analysis, and Segment Compare to uncover hidden insights that might take days to discover otherwise
  • Analytics reports are easy to download, email, and share with colleagues and stakeholders
  • Data is also available through Workspace, Ad Hoc Analysis, Report Builder, and Data Warehouse
I know, you're excited! But before you kick off your next Target activity using Analytics as the data source, there are a few things you should do to be prepared.
  • Make sure you have the proper permissions to both configure your Target activity & view the data using Analytics. More on that here.
  • Get to know your Analytics variable map.  If the Analytics & Target users at your company are not the same people, schedule a collaborative session to understand what data is (and isn't) being collected and what each variable means.
  • Learn the rules. In Analytics variable allocation & expiration may be different than Target. Get familiar with the nuances so you can correctly interpret your Target activity performance. More on that here and here. The good news is, nearly all activity types are supported with A4T.
  • Partner with your Analytics power users. Agree upon how Target activity results will be analyzed & shared.

If you're ready to get started and you prefer a video walk-through of the following content, check out a previously recorded webinar entitled Analytics for Target: The smarter way to optimize.

I just launched a Target activity using Analytics for measurement. Where & how should I perform my analysis?

For starters, review aggregate performance to determine which experience is the overall winner. While either the Analytics interface or the Target interface will show the same data, Adobe recommends using the Analytics interface for analysis.

If you start in Adobe Target simply click "View Report in Analytics" to switch interfaces.

Once Analytics, you'll land in the Reports interface. Analytics automatically detects the start & end dates of your Target activity. You can also navigate directly to this out-of-the-box view by typing Target Activities in the Analytics navigation and then drilling into the activity name to see experience level reporting.

Important:

  • The report date range in Analytics matters because the tnt variable that collects A4T data is a list variable that persists for 90 days with full allocation. This means conversions will continue to be attributed to your activity even after it is disabled.
  • Also, note that the default cookie expiration for Target is 14 days. If your Target cookie has a duration set to something that is different than the tnt variable expiration in Analytics (default 90 days), it's a good idea to align them. For assistance, contact Adobe Client Care to have them set your Analytics tnt variable to match up with your Target cookie duration.

While Reports & Analytics will continue to be available, the interface of choice going forward is Analysis Workspace.

Workspace is my favorite! I heard there's a new panel. How does it work?

You heard correct! Adobe released a brand spanking new panel in Analysis Workspace built specifically for Target activities. To get started, launch Analysis Workspace and click Create a New Project. Drag the Analytics for Target panel into the Workspace canvas and you're ready to go. Check out Adobe Help for a quick overview of panel inputs.

Which metrics do I choose?
  • Success Metrics: You can choose up to 3 metrics at a time. Yes 3! The panel will calculate lift, confidence, and conversion trending for each metric. Use any conversion event already being collected in Analytics or select the Activity Conversion metric (a.k.a. the Target Goal Metric). Calculated metrics can be applied to the panel after build but are not supported with lift & confidence.
  • Normalizing Metric: For activity types that are assigned and persist at the visitor level (A/B tests, MVT, Recommendations, Auto Allocate) the ideal normalizing metric is Visitors. If you use the Visits metric, know that it reports on each session that a visitor initiates after they qualify for the experience, regardless of whether they saw the test experience in subsequent visits.
This panel is fancy! How do I interpret what I'm seeing?

Freeform table

Tip: You can apply breakdowns such as browser, device type, or marketing channel to the table but note that lift & confidence will disappear. Keep lift & confidence with a segment or dimension drop down at the top of the panel instead.
  • Target experiences within the chosen activity are listed in each row. The control experience is bold.
  • The normalizing metric (typically Visitors) is listed in the first column, and the first of three success metrics is listed next to that. Each success metric has its own table.
  • The conversion rate is calculated by dividing the success metric by the normalizing metric (Visitors).
  • Lift is broken out into three columns: Lower, Mid, and Upper. The expected lift is actually a range that narrows as confidence increases. The Lower and Upper lift are the boundaries of that range while the Mid value is the average of the two (also shown in Reports & Analytics). Note that lift is a locked metric and it cannot be moved, deleted, edited, or broken down.
  • Confidence is calculated against the measured lift and is an indication of whether the results are statistically significant and therefore reproduce-able. Conditional formatting is applied to the confidence metric to highlight when an experience approaches the desired confidence threshold (95%+). Confidence is also a locked metric.
  • Apply a segment either at the panel level or within the freeform table. Note that if you add it within the freeform table you must overlay it across the entire table to preserve the lift & confidence calculations. Column level segments are not supported at this time.

The same rules of engagement govern results interpretation in the Analytics for Target panel. For a review of test analysis best practices refer to Adobe Help.

Line graph

The line visualization complements the freeform table by displaying a view of conversion over time for each experience. The trended conversion rate highlights when the data normalizes and which experience consistently converts the best.

Tip: Click the gear icon in the top right of the line visualization to change the granularity to Day. This will display more specific detail around conversion trending.

To revise panel inputs or choose different metrics, click the pencil in the top right of the panel.

The Analytics for Target panel is a jump start for performing top level analysis with lift & confidence in Analysis Workspace. Within the panel you can add additional metrics, apply segments or breakdowns, and create new visualizations.

What else can I do in Analysis Workspace?

Go deeper in Analysis Workspace to visualize the data or uncover insights hidden beneath the surface.

A critical step to using Analysis Workspace to understand Target activity performance is building an audience for visitors that were exposed to each experience. Create a saved segment like the example below, or simply use the Target Activity>Experience dimension for a temporary segment.

Tip: Adobe recommends using a HIT container to include behavior only after qualifying for an activity experience. Using a Visit or Visitor container expands the scope of the audience to include actions that happened prior to test qualification.

Apply an experience segment to a Flow visualization to examine whether your activity treatments fundamentally change how users navigate the site.

Tip: To easily toggle between experiences use a segment drop-down. Search for the Target Activity>Experience dimension and multi-select the experiences in your Activity. When you drag them to the canvas, hit Shift to create a drop-down.

Compare each of your activity audiences side-by-side in a Fallout visualization to see how fallout differs downstream.

Leverage the Segment Compare panel to quickly discover differences between visitors who were exposed to the control experiences vs. another experience.

Tip: Segment Compare allows you to compare in depth only 2 segments at a time. Create multiple panels to compare experiences against the control, or even against each other.

For more on Analysis Workspace, check out our tips & tricks guide for Workspace ninjas.

Word on the street is A4T works with Auto Target and Auto Allocate now. Is it true?

Heck yes! Analytics for Target works with almost every activity type. When it comes time to configure and later analyze performance of your Auto Target activity, there are some nuances to be aware of. See Experience League for a step-by-step tutorial to build out reporting for these automated activity types. Also, see the YouTube video for using an A4T Panel with an Auto-Target test.

For guidance on implementation of Analytics for Target (A4T), please reference the following materials:

Congratulations! You made it to the end of this tutorial.

Now that you're a pro at performing Target activity analysis in Analytics, if you have improvements you'd like to see incorporated into future product releases, please contribute by submitting a suggestion or voting for existing ideas on our Adobe Experience League.

Here are a few fan favorites:

Ready for more content?

Visit adobe.ly/aaresources for a full list of Adobe Analytics Spark pages & other helpful resources.

Created By
Kaela Cusack
Appreciate

Credits:

Created with an image by Karolina Szczur - "untitled image"