Resources

Technology is a vitally important part of testing and optimization, but it’s only part of the solution. Testing and optimization projects need planning, execution, and reporting to reach their full potential.

Getting started with site optimization

A testing and site optimization campaign is as only as good as its design. Widemile has developed the expertise and Best Practices to get the most out of every optimization campaign. To help get our users started, collected here are some of the top optimization blog posts from one of Widemile’s expert optimization analysts:

• Why testing should be a part of your website and marketing toolset
• The basics of multivariate testing
• How to execute effective A/B Split tests
• An easy 3 step optimization plan

For more recommended articles, check out a Lesson Guide on Billy's Blog or go to his Home Page. You also can keep up by subscribing to the Billy’s Blog RSS feed or following him on Twitter.

Optimization testing

Widemile Optimize™ supports all major test types: Baseline, A/B Split, and two forms of Multivariate; Full and Fractional Factorial.

Baseline Analytics

Widemile Optimize collects essential analytics to validate web traffic and conversion rate data. Baseline analytics are best run at the beginning of a project to establish accurate numbers for comparison later on in the testing and optimization lifecycle.

A/B Split Tests

Also referred to as Split tests, this test type is used to evaluate two or more different web pages. Split tests are useful for determining the better of two or more vastly different web page design approaches such as top navigation versus left-side navigation or Flash–based design versus static HTML.

Multivariate Tests

These tests are the core of Widemile Optimize and are for testing different content including copy, images, rich–media, and nearly any other type of content to determine the most effective combination of that content for your web traffic. All content variations are managed by the marketer and their web team with no interference or hosting from Widemile. Once a test is running, Widemile Optimize simply does the rest; it records and tracks web visits and makes reporting on this data easy and straight–forward allowing the marketer to provide data that proves one set of content is most effective.

There are two methods available for a multivariate test:

1. Fractional Factorial

   This method of multivariate testing uses Widemile’s patent-pending Design of Experiments methodology to produce highly accurate results as quickly as possible using balanced orthogonal arrays and advanced statistical modeling. Multivariate tests using Widemile’s fractional factorial method actionable data in a fraction of the time required by an equivalent test using the full factorial method.

2. Full Factorial

   This method of multivariate testing evaluates all possible experimental combinations and interactions. This method is useful when there are very few factors to consider in the content equation. Full factorial tests are typically many times slower to provide actionable data than fractional factorial but are useful in some cases.

Back to Top