Do Your Campaign Tests Pass the Test?
Retail Marketer's Guide to A/B Testing Success
A retailer’s main goal should be to increase the three factors of a sale – website traffic, average order values (AOV), and conversion rates. These three factors are key business metrics used to determine the success of a website, campaign, advertisement, or other marketing initiative. Each and every marketing initiative can intentionally and/or inadvertently affect conversion rates.
Conversions result from sales, offers, clicks, purchases, downloads, applications, memberships, or any other measurable action. On average, conversion rates range from 1% to 3%. That means as much as 99% of a site’s visitors aren’t answering the call to action that is being measured.
So why are conversion rates so low? Our experience suggests it’s because improving conversion is a complex challenge.
Coming from the perspective of a business that works with clients to improve their website conversions, the best conversion optimization practices are not just about running a large number of tests and reporting on the results. Rather, it involves using data to gain valuable insight and only then running the tests that really count. This is why it’s essential to test thoroughly and target accordingly. By using the correct testing and targeting methods, a retailer can see an increase in conversion rates of 300% or more.
A/B Testing Roadmap
When planning to perform an A/B Test, we recommend that you prepare a simple testing roadmap. An A/B Testing Roadmap will lay out those campaigns you plan to run, why you’re choosing to run these campaigns, and what you’re hoping to learn from the campaign. From here, you can easily track and monitor the results of the campaigns.
We recommend you identify the following items:
TEST NAME — or an Identification Number, any way to identify what you’re testing
LOCATION — Where on your site will the test be performed?
START DATE — Make sure your start date does not overlap with other test dates
DESCRIPTION — Describe what the test will entail
HYPOTHESIS AND/OR QUESTION — Determine what effect and/or result you think the A/B test will have
IMPACT — Create a score for the impact you feel this test will have based on a 1-3 scale
EFFORT — Create a score for the effort needed to implement the campaign based on the same 1-3 scale
And remember, every marketer should create a schedule of A/B test ideas, and track the results so optimization is always moving forward.
So we are agreed that A/B Testing is an effective and proven way to increase average order values, boost conversion rates, and drive more sales. But with so many variable, it can be hard to decide exactly what to test.
Here are some ideas we think will help you get the most out of your campaigns.
HEADLINE or product description (size, color, wording)
The DEAL or PROMOTION being offered (content, working, placement)
The CALL TO ACTION button (e.g. the button’s wording, size, color, and placement)
SOCIAL MEDIA buttons (size, placement, and wording)
CREATIVE for display ads and landing or product pages
AMOUNT OF TEXT on the page (e.g. short phrases vs. long descriptions)
The exact AUDIENCE SEGMENTATION to be targeted
If you are optimizing ads or offers for a new product or special, you may also want to test the existing ads/offers against new ads/offers containing all of the “best practices” you have discovered for designing landing pages, order pages, special promotions, etc.
And of course, don’t forget to keep track of all the results. Keep in mind that test results are conditional to the time frame in which you run the test and should be reevaluated at some point in the future, typically within one year.
Organizations that make testing part of their culture reap true rewards. To see significant lifts in key business metrics like lead generation, sales, and conversion rates, testing must be an ongoing practice since the marketplace is ever-changing — as are consumers.
Sign up for SteelHouse updates
Learn how SteelHouse can help you drive new visitors to your site and bring them back to convert — all from a single platform.