• Internet
• 09.05.2017
• EN

OpenText™ | Web Site Management How to use Optimost Testing with Web Site Management

White Paper

How to use Optimost Testing with Web Site Management

Product: OpenText Web Site Management and Delivery Server
Version: 16.0.1
Task/Topic: Features and Functionality, Deployment, Administration​
Audience: Administrators, Decision Makers
Platform: All
Document ID: 510016
Author: Henning Hinrichs, Principal Software Architect, OpenText Web Site Management
Updated: May 9, 2017
Contents:

Overview: Testing web site objectives

Management Summary
This article discusses the use of A/B testing to better reach objectives with web sites. It presents simple example showcases implementing A/B testing using OpenText Optimost Visual Test with OpenText Web Site Management (WSM). The article ends with a section discussing further possibilities and a chapter answering some frequently asked questions.

Why A/B testing
A/B testing is used to increase the so-called conversion rate, meaning to increase transactions per visit. However, A/B testing can be used more generally to test hypotheses to better reaching objectives.

Web analytics does not know your objectives
You might want to know which users do what on your web site. To this end, you need web analytics. However, analytics just collects and presents data as-is. Analytics does not know about the objectives you want to pursue with your web site. Web analytics cannot answer questions like whether you should have done something different to better achieve your objectives. For example, more clicks, orders, downloads per visit.

Use opinions and empirical data to discuss variants and audiences
This is where variants enter the scene. You can discuss pros and cons of variants of your web pages and the hypotheses backing them with as many peers and customers you can find. Typically, there will be many opinions, preferences, reasons, and hope to select the best variant for live production. You may even decide to produce different variants for different audiences. Arguments, opinions and even speculations are required and valuable to understand and classify web site usage and user motivation.
However, real users with their concrete interest in real-life contexts may behave unexpectedly. You only know if you get real data from them, by presenting different variants of one or more pages to your user base - and then use analytics to find out which variant does better in reaching defined objectives. With this so-called A/B testing, you can check hypotheses on web site effectiveness by empirical data, turning opinion into knowledge and often producing significant enhancements.

A/B testing process and tools
As simple as the A/B testing concept is, it nevertheless requires some steps to implement. For some of these steps, you will need tools. Typically, the IT department or technical personas do not define objectives for a web site. Hence, the tools should be accessible by personas with as little technical assistance as possible.
Since A/B testing variants are (during the testing period) presented live to users of the web site, a partial or similar process to the releasing of web site changes should be followed.

...

Das komplette Dokument ist im OpenText™ | My Support hinterlegt.