The following section describes different use cases for testing using ATG Campaign Optimizer. Each use case includes a link to the section describing exactly how to perform the described functionality. Use cases that have (Commerce Only) in the title are available only to ATG Campaign Optimizer for Commerce customers.
Testing Slot Images
A business user wants to run a three-day test to determine whether the image of a female model or a male model works best on the home page.
The business user sets up a three day test where 40% of the site’s visitors will see the female model image, 40% of the site’s visitors will see the male model image, and 20% of the site’s visitors will see the image that is live on the site today. She selects the slot that she wants to test from a list of slots that are live on the site. As she sets up the test, she is able to preview the two test images as the images will appear to test participants.
See Testing Repository Content for details on setting up this type of test.
Testing a Promotion (Commerce Only)
A marketing manager is planning a large promotion for the Back-to-School season. She has to choose between a “$10 off any order over $100” and a “$20 off any order over $200” offer. She decides that she’ll run a one week test on the web site to see which offer is more profitable.
There are four test groups that she wants to measure. The first is a group that will receive the Purple Ad and the $10 off promotion, the second will receive the Blue Ad and the $10 off promotion, the third will see the Purple Ad and the $20 off promotion and the fourth will see the Blue Ad and the $20 off promotion. The control group will receive no ad and no promotion.
She sets up a seven-day test where each group receives 20% of the traffic and the control group receives the remaining 20%. She picks her promotions and images from a list of promotions and a list of content pieces.
See Testing Promotions and Testing Repository Content for details on setting up this type of test.
Changing Price Location (Commerce Only)
A business user speculates that conversion will increase for high price point items if the Price Block is moved below the Product Description. The business user asks for page designers to create a new version of the Product Detail Page template in which the price block is below the description.
The business user sets up a test to determine which template performs better. The new template is used for all Armoires and Couches, but the old template is used for all other types of product. The user sets up the test so that 50% of the site’s visitors will see the new template in those two categories and 50% will see the old template. He sets the test to run for seven days.
SeeTesting Product Templates for details on setting up this type of test.
Testing Category Design (Commerce Only)
A site administrator wants to experiment with the category template used for Queen Anne furniture. Currently, the catalog uses one category template for all Queen Anne furniture categories. He would like to test the change in customer performance between the new category template and the old one, by providing the new template to the Queen Anne chairs and Queen Anne beds categories and the old template to all other Queen Anne categories, such as Queen Anne tables.
He creates a category template that uses a new color scheme and changes the placement of the components on the page. He sets up a two-day test where 50% of the site visitors see the test template for Queen Anne chairs and Queen Anne beds. The remaining 50% see the original template.
See Testing Category Templates for details on setting up this type of test.
Testing Product Recommendation Strategies (Commerce Only)
A programmer has created a product affinity engine that can make product recommendations based on other people’s purchase habits. She wants to prove that the content that is served by her engine will generate more in sales than the current process of fixed content (related products). The programmer decides to create a new product template that incorporates her recommendations in the ‘related product’ slots. She creates a two-day test that reveals the new template to 50% of the site’s visitors, the old template to 50% of the site’s visitors.
See Testing Product Templates for details on setting up this type of test.
Combining Multiple Tests
A business user would like to understand the effect of a new site color scheme on purchasing behavior. He would like to initially test the Product Templates and the Category Templates without switching the entire site over. He sets up a test where 50% of the site visitors will receive the new color version of the Product Template and the Category Template. The remaining 50% will see the old (current) color scheme.
See Testing Product Templates and Testing Category Templates for details on setting up this type of test.
Modifying a Running Test
A business user is reviewing her ATG Campaign Optimizer test results after the first day of testing. She notices that she is not gathering enough data, so she decides to extend the test for 5 more days. She can alter the end date for the test from the 10th to the 15th.
SeeEditing Tests for more information.
Copying an Old Test
A business user would like to create a new test campaign for the same slot that he recently ran another test on. He opens the old test and copies it with a new test name. He alters the content that is designated for the slot to his new content and alters the dates for the future.
See Copying Tests for more information.
Reviewing Results of a Running Test
A business user is curious about the first day’s results from a test that she set up. She looks at her ongoing test’s interim results on the test’s Results page.
See Viewing Test Results for more information.
Learning About Missed Revenue Opportunities (Commerce Only)
A business user notices a drop in product sales. He wonders how many users exhibit interest in purchasing products, by adding them to a shopping cart, but don’t actually purchase them. Using the Results page, he can see how much potential revenue is represented by orders abandoned by members and guests. If he finds that a significant amount of revenue is lost to abandoned orders, he may create other tests to determine when, during the purchase process, abandonment most often occurs.
See Viewing Test Results for more information.
