RSS feed Get our RSS feed

News by Topic

BizReport : Ecommerce archives : December 13, 2013

How to create a solid A/B testing strategy

One of the greatest challenges for businesses is getting the data right. Online retailers should take prudent effort to make sure the infrastructure and process they've implemented is sound. For example, the "novelty effect" suggests that just because a change has an impact initially, doesn't mean it can be sustained over time. Will your customers grow tired of cyber deals every ten minutes for a week? Testing is one way to determine what customers really want.

by Kristina Knight

Kristina: What are some common mistakes companies make in A/B Testing

Dan Siroker, CEO of Optimizely: One common mistake is not asking the right questions. Consider first what you want the answers to be. Those hypotheses will then help you decide what to measure. Common examples of tests for online retail include: homepage bounce rates, category page views, product page views, shopping cart ads, and all stages in a checkout flow all the way to the 'Thank You' page. In general, rather than asking "What are the variations we are testing?" consider asking, "What question are we trying to answer?" as a means to get more effective and relevant results.

Another is not making the web site call to action loud and clear. As the site-wide checkout entry point, your primary calls to action--like "add to cart" or "sign up for emails"--are critically important to test. These tests should clarify what precisely you want your website to accomplish and whether you are effectively directing shoppers to accomplish those goals.

Kristina: How can marketers get started with A/B testing?

Dan: To ensure the best return on your effort, first look at your web analytics to see which pages have the most room for improvement. You'll want to first attack these pages. Next look at the traffic volume that hits each page and determine which are the most high-cost, such as those that originate from paid search or affiliate programs. Lastly, assess the ease of implementation--from both a cost and technical perspective. Once you've conducted this type of assessment, you can rank the pages and start the A/B testing where you need it most. You'll quickly find that the tests take the guesswork out of website optimization and enable data-backed decisions that shift business conversations from "we think" to "we know."

You can read part one of my chat with Dan, including his top five hubs to test, here.

Tags: a/b testing, ecommerce tips, Optimizely, website performance

Subscribe to BizReport



Copyright © 1999- BizReport. All rights reserved.
Republication or redistribution of BizReport content is expressly prohibited without the prior written consent.
BizReport shall not be liable for any errors in the content, or for any actions taken in reliance thereon.