Identify the best offer using A/B Testing in Unica Interact

Every marketer aims to achieve better click rates, easy conversions, reduce the churn rate, and create a memorable experience for their customers. To accomplish it, the decisions they take while sending out emails, designing CTAs, landing pages, and even sending out offers should be inline with the customer they are targeting. To avoid the assumptions of what shall work and what won’t, A/B testing, also known as split testing, is the marketing technique that utilizes two versions of the webpage, application, offers, email, or any other marketing asset to analyze and understand which one performs better.

A/B testing is essential as different customers may react differently to varied situations. For example, some might want the ‘Buy Now’ button in the center of the page, and others might like on the bottom right corner. But which one of them leads to the most conversions? Is the answer that A/B testing can provide.

How A/B Testing works in Interact?

In Interact, an A/B Testing is a test done for a treatment rule with all its properties fixed, but only one parameter varies. In 12.1, this parameter is a combination of offer and its offer attributes. In an A/B testing of a rule, each varying parameter represents one branch with audience distribution percentage. Depending on the response of the Audience for an Offer in a set Test Period, you can choose the best suitable offer for that treatment rule.

Set up A/B Testing for Treatment Rule

• A new option A/B, is added to the Rule menu; you can directly land to A/B Testing pane by using it.

• On A/B Testing tab, you can enable A/B Test for a Rule by selecting Enabled Check Box.
• After enabling the A/B Test for a Rule, a Default branch will be added by copying details from the base Rule.

• You can even add more branches by clicking ‘Add Branch.’

• You can change the Branch name, which should be unique within A/B Testing.
• A branch can be enabled and disabled. Only enabled branches will participate in A/B Testing at runtime.
• Sampling % is the distribution of the Audience between branches.
• Also, you can add Expression to Offer Attributes. After adding an expression to offer attributes, its count will be displayed in the ‘Offer Attributes’ column, and a dropdown will appear to check more details.

• You can also set the ‘Test Period’ for which A/B Test can be enabled.
• A Default branch is a branch out of A/B testing branches, which is executed once the testing completes, i.e., after its expiration date. It can be completely independent of base rule for cell code, offer, or parametrized offer attributes. For the Unica 12.1 release, you need to designate a default branch, and it must always be enabled. In the upcoming release, the selection of the Default branch will automatically depend on Branch performance.
• After saving A/B Testing details on the rule list page, an Orange ‘A/B Testing’ icon will appear for a Rule.

• You can filter out only A/B test Rule by Selecting the‘ Show A/B testing rules only’ check box on Rule list Edit or View mode.

• After enabling ‘Show A/B testing rules only,’ filter you will get a limited set of Columns and can see Branches details like Sampling% and number of Offer Attributes with expression by expanding section.

• Also, with the ‘Show A/B testing rules only’ filter, you can further filter on rules by using a search by Segment or Zone.
• Once Interactive Channel with A/B Testing is deployed, A/B Testing details are also displayed for rules on deploymentInfo.jsp.

• To save A/B Test data in Systems tables, two new tables were added in Design Time(Campaign) Schema UACI_ABTest and UACI_ABTestBranch.

Analyzing A/B Testing Results

You can analyze branch performance and set that branch’s offer as the default offer for treatment rule in the following ways.

1. To check Branch performance, the Customer needs to add a column “ABTestBranch” of string data type into the target CHStaging and DTLContactHist tables, and then the customer can run a query or create Custom reports to determine branch performance and then set the default branch for the rule, in the upcoming release we are going to add new reports to show the performance of all A/B testing, organized by interactive channel and then strategy.

2. Customer can configure Cognos, or Birt reports and check ‘Interactive Offer Performance Over Time’ report for Offer performance summary and decide which branch offer is doing well.

The ability to conduct A/B Testing with Unica Interact will help customers identify the best offer for their Treatment Rules using Comparative Analysis of Offers. This will help them in making more customer-focused decisions and ultimately increasing the conversion rate of the offer. You can reach out to us for any more queries, and we will be happy to help.