The Audience Segmentation Cloudlet provides hassle-free traffic segmentation and session stickiness without degrading performance. This is often beneficial for A/B and multivariate testing.

Once you activate an Audience Segmentation policy, it identifies an inbound request and applies the rules you defined to that request. Then, at the edge, the Cloudlet creates stable test populations by assigning a cookie value to the user. Audience Segmentation can be used with your Akamai service to provide the desired experience for that user.

Workflow

Audience Segmentation identifies the inbound request and determines the appropriate segment for that request.

The edge server then assigns the session cookie to the request. Audience Segmentation uses this cookie to randomly assign users to a population and to maintain sticky populations.

Here's a diagram of how this works:

AS_workflow

Use cases

You can use Audience Segmentation to help with:

Item	Audience Segmentation Use Case
Testing	Use Audience Segmentation to help with both A/B and multivariate testing. A/B testing lets you test changes to a page by sending some users to the new page and the others to the current page. Based on the data collected, you can see whether the response to the new page is positive. Multivariate testing helps you determine the web asset variations that perform the best.
Release activities	With Audience Segmentation, you can set up different traffic segments to more slowly roll out new functionality to avoid any potential release issues as well as degradation in the user experience.
Layered user experience	Use Audience Segmentation to provide a slightly different look and feel for your special customers or to support promotions.

To learn more about Cloudlets basics, see:
How Cloudlets work
Cloudlets terminology
Property and Cloudlet policy considerations