Process: Monitor marketing experiment
Flow
Objective
The objective of monitoring marketing experiment is to determine the most effective site content. The experiment results are compared against corresponding result for the control content using such factors as acquisition of new customers, attracting repeat visits, and improved conversion rates.
Description
To perform the analysis of how the experiment compares relative to the original content, the experiments system uses its own set of listeners and database tables to keep separate statistics, which are available to validate the experiment. Based on the statistical feedback available through the WebSphere Commerce Accelerator, the Marketing Manager can immediately choose to update the original control element to incorporate the experiment's test element. For example, if an experiment is testing advertising content, and the Marketing Manager finds that the experiment element has a conversion rate that doubles the control element's conversion rate, they can update the control element such that it uses the content defined in the test element.
Features
After the statistical data is reviewed, the most effective element can be converted to the control content and scheduled to go live in the store front. The statistical report contains the following data:
- number of views
- number of clicks
- ratio of number of views to number of clicks
- number of orders
- ratio of displays or clicks on experimental content to a placed order
- total and average order amounts
Customization
- More statistical data can be added to the report to provide more assistant on determining the most effective element in the experiment.
Edition
Professional, Enterprise
Tasks
| Task | Description | Role |
|---|---|---|
| Change Web activity |
Depending on the nature of the change, one or more business processes described earlier may be used to change the content shown through the Web activity. The following list shows an example of possible changes:
|
Marketing Manager |
| Copy and Change Experiment |
After the experiment is completed, the marketer wants to change the definition of the experiment in order to improve the accuracy of statistics for a better insight of how well the control Content works. The marketer will copy the experiment, and change its definition. |
Marketing Manager |
| Review experiment statistics |
After the marketing experiment ends, the clickstream and order statistics related to the experiment can be reviewed, and an action is determined to either change the control Web activity to make it more effective, copy the experiment and change its definition to improve the accuracy of statistics to give the marketer a better insight of how well the control Content works, or there is no need to take any further action if the experiment shows that the control element is actually more effective than the test elements. |
Marketing Manager |
Business artifacts
(C) Copyright IBM Corporation 1996, 2006. All Rights Reserved.