Process: Marketing Experimentation
Flow
Objective
The objective of marketing experimentation is to find the best method of maximizing marketing campaign effectiveness by testing alternative versions of site content. Campaign effectiveness is measured by such factors as acquisition of new customers, attracting repeat visits, and improved conversion rates.
Description
Marketing Managers can use Marketing Experiments (sometimes referred to as A/B testing) to test alternate site content as a means to maximize revenues. Existing marketing activities (control elements) may be tested with alternate experimental elements. The experimental elements include testing of alternate promotions, content, images, customer segments, or e-Marketing spot locations. The Marketing Manager can then compare the results from these experiments to determine the effectiveness of each specific change and the impact it has on customer behavior. Statistical information is accumulated to generate reports for each experiment. Once useful results are obtained, business users can make informed decisions and update the original activity to reflect the optimal content to maximize revenues.
During the experiment's creation, the marketing manager specifies a display ratio, which determines the percentage of customers who will see the experiment (or experiments) and the percentage that will see the control element. When a customer accesses the site, he will be randomly assigned to either the control or any of the experimental elements.
Features
The experiment can be set up with one of 3 different content control options: 1) each time the e-Marketing spot is visited, the content to display to the shopper will be re-evaluated in a random nature; 2) each time the e-Marketing spot is visited in the same shopping session, the shopper will see the same content as initially presented to the shopper (experimental or control); or 3) each time the e-Marketing spot is visited within the same shopping session or upon subsequent shopping sessions (for example, close the browser and open new browser and go to the store the next day), the shopper will see the same content as intially shown (experimental or control).This will guarantee a consistent browsing experience and more consistent experiment results. This last option requires the persistent session feature to be enabled so that the experiment information for a shopper who has visited the site while the experiment is active will be saved and used in subsequent visits to the site.
The following types of elements can be tested against a control web activity:
- New Web activity displayed in a particular e-Marketing Spot
- New content displayed in support of an activity
- Alternate target customer segment
- Alternate e-Marketing Spot in which the activity displays
Customization
- New types of testing can be supported for the e-Marketing Spot store element type.
- New types of store element can be introduced such as store page template and style.
Edition
Professional, Enterprise
Subprocesses
Business artifacts
(C) Copyright IBM Corporation 1996, 2006. All Rights Reserved.