Optimizely Blog

Tips & Tricks for Building Your Experimentation Program

X

Download our FREE Testing Toolkit for A/B testing ideas, planning worksheets, presentation templates, and more!

Get It Now




We’re glad to welcome Francis Teo, CEO and lead consultant at Conversions Guaranteed, to the Optimizely blog for today’s guest post.

What is a Multi-Page Test?

In the world of A/B testing, a multi-page test refers to an A/B test run simultaneously across multiple pages of your website.

In a multi-page test, you can test changes (what we call “variations”) across different pages of your site. On an e-commerce site, for example, you can test variations of your site-wide navigation across the home page, the category pages, and the cart pages, all in the same experiment.

Screen Shot 2013-06-11 at 6.27.42 PM

Sometimes knowing when to run a multi-page test can be a challenge. Here are some great examples to get your started.

When to Use a Multi-Page Test?

1. When you want to provide a consistent user experience: Let’s say you’re testing a new header graphic. A multi-page experiment will ensure the correct header is consistently shown to visitors across your website. If a visitor is bucketed into the original variation, they will always see the control header graphic; if a visitor is bucketed into the new header variation, they will always see the new header.

Having a consistent user experience during a test is important – you don’t want to cause mistrust or confusion among users by showing changing interfaces throughout your website.

2. When you’re redesigning your website: Thinking about rehauling your website? Use a multi-page experiment to see how the new site design performs against to the original. E-commerce retailer Finish Line Inc. rolled out a new site last year without doing any A/B testing. The new site performed significantly worse than expected, causing Finish Line to lose $3 million in revenue before reverting back to the original design. Running a multi-page experiment on your redesigned site – perhaps on a smaller portion of site traffic – can give you a better idea of how the new site will perform before rolling it out completely.

3. When you’re not Amazon or Google: For sites with lower traffic, a multi-page experiment can, in some cases, reduce the time it takes to reach a statistically valid test. Testing more elements on more pages is likely to produce a larger conversion variance, as compared to just changing one element.

Quicker test results will help you build your customer model more quickly, as you learn what works (or does not work) for your users.

4. When you want to improve your conversion funnel: Do customers prefer a 2-step or a 5-step checkout funnel? Are they more willing to input more information up-front, or should those extra form fields be moved further down the funnel? For e-commerce sites, revealing the right information at the right point in the funnel can boost conversions significantly.

Test for Success

Multi-page testing is one of many strategies that can be deployed using Optimizely as a split testing tool. As part of a broader testing strategy, informed by data and customer insights, multi-page testing can help achieve significant wins and long-term results.

Optimizely X
comments powered by Disqus