A BLOG FOR E-COMMERCE LEADERS

SLI Blog

Posts with A/B Testing Tag

Find Your Way Out of Any Paper Jam with This Online Store

Posted by Dawn MortensenConferences, eCommerce, Navigation, Site Search, Success Stories1 comment

This is third in a series of success stories on eCommerce sites that are growing their business with advanced site search. Also in this series, read about Party Supplies Delivered and Vermont Teddy Bear, and see live presentations of these stories at the SLI Theater at IRCE June 3-6.

Jam Paper & Envelope, a top source for paper crafts, creative giftwrap and all other things paper-related, has been featured in Real Simple, Martha Stewart Living and Country Living Magazine, among others. With its flagship New York City store, a successful online retail operation, and two other store locations, the paper company has been able to succeed both in the brick-and-mortar and eCommerce worlds.

JamPaper.com offers more than a million variations of specialty envelopes, paper, cardstock, gift cards and other paper goods, with 70% of its sales going to businesses that place large orders for marketing promotions, weddings or other events. The business started two generations ago and has remained in the family, with grandson Andrew Jacobs leading the effort to create the eCommerce site. Jam Paper first went online in 2004 and the site traffic is growing steadily.

“Our initial site search was rudimentary,” Jacobs said. “But since we’ve started online we’ve been constantly working on improving.”

One of the challenges for JamPaper.com was to help their users find the exact product they wanted among the enormous number of items and pages. Envelopes, for example, are one of their most common searches, but envelopes can come in 50 colors and 50 sizes – with many small variations for each. Since they improved their site search and navigation, Jam Paper has made the following improvements so users can more easily find and purchase the exact product(s) they want:

  • Made site search results more relevant and incorporated sales data into search results tuning so the most popular items come up higher in results.
  • Added synonym suggestions to improve the results on search terms that previously turned up with poor results.
  • Incorporated Rich Auto Complete to auto-populate suggestions based on popular search terms and take users directly to the product page.
  • Analyzed and restructured data internally to improve search results.
  • Added merchandising banners to search results pages.
  • Used A/B testing to identify additional improvements that will increase conversions.

By analyzing their search data on an ongoing basis, Jam Paper continuously improves their online user experience. The impact on online sales is “crawling upwards and keeps getting better,” Jacobs said.

At IRCE in Chicago June 4 and 5, Andrew Jacobs will be at the SLI Theater to share the tips and tricks he’s learned for improving Jam Paper’s eCommerce site search, navigation and SEO. Stop by Booth #501 to meet him or to get a free site critique from the SLI Systems team.

Other SLI Theater presenters include Steiner Tractor, Envelopes.com, Samuels Jewelers and ReNew Life. See the full schedule of presentations here.

A/B Testing and Why You Should Do It

Posted by Kemberly GongeCommerce2 comments

A/B testing is one of the most important things you should be doing for your eCommerce site. In eCommerce, your site’s layout, design and feature set are a crucial part of the product – not just the physical products you sell. Yet many businesses don’t place high importance on testing. Without it, you’re essentially displaying products without ever knowing how your customers respond to your layouts.

A/B tests compare a variable against a control to determine which one is more popular. For eCommerce sites, elements of control page “A” are compared to the new version “B” to find which one increases interest in the page. This could be in the form of click-through rates, conversions, pages viewed per visit, or whatever goal is specified.

Say you want to encourage more people to click a button on your page to sign up for a service. To test the effectiveness of the button, you’d use the current page/ button design as your control and a test it against a new design. Both designs would be released simultaneously to a customer sample, and their clicks during the test period would be tallied to see which button drove more clicks. You’ll usually see a pretty clear indication of one particular design that users prefer. A/B testing is a simple, yet effective way to determine which version “wins”.

With the results of individual tests and the collective results over time, testers will get actionable data about what design, layout and features users prefer.

But repeat after me: “one design does not fit all”. User demographics and preference play heavily into what works for one site but not for another. Even though your site may sell something similar to another, your audience may be different and users may behave in very dissimilar, contradictory ways. It’s important to test your site specifically, and not necessarily follow others in your industry.

Instead of grasping at straws when deciding on design or feature changes on your eCommerce site pages, consider A/B testing to help give clearer, data-driven insights that can provide concrete recommendations to improve your site’s performance. The only thing you should not A/B test is the decision to A/B test.

Have you been shocked by results from A/B tests? We’d love to hear what you learned.