This is only a test.
Great post from the experts at the Email Experience Council and our friends at Marketing Profs. Testing is the key to getting results. The concept they push here, A/B split testing, is one of the latest features of Online Outbox. Split testing allows a marketer to get a true litmus of the effectiveness of a campaign. Here goes:
“A great way to capitalize on the democratic medium of email is to put your burning questions, late-night hunches, and out-of-the-box ideas to the test with an A/B split test,” says Megan Walsh in a post at the Email Experience Council blog. Here’s some of her advice for using split testing to try out new ideas and boost results:
Focus on one variable at a time. When testing subject lines, for instance, you want to know how many subscribers opened the message. Clicks and conversions, meanwhile, matter when you’re testing calls to action. Remember that a change in one variable can have an impact on another. Always be clear about what you’re trying to test/achieve, Walsh advises.
Use a random distribution for A and B audience groups. “The sizes of the segments don’t need to be the same if the key metric you are looking to influence is expressed as a ‘rate,’” she says, “but they do need to have the same general characteristics to be a fair test.”
Take it to the next level. As you become a testing junkie, begin to explore how different segments behave. Williams-Sonoma knew that including a featured product’s price on the hero image boosted clicks and conversions. The company recently discovered, however, that the customers driving this response were those who spent more than $100 on an average purchase; customers who spent less were more likely to click on an image without a price.
The Po!nt: Never stop testing. Something that works today won’t necessarily work tomorrow: customer preferences constantly vary.
One advantage of using Split Testing with Online Outbox is that you’re not limited to a single split. In other words, you’re not limited to A/B — you could segment and test A/B/C/D… or more, but you get the idea… Get more info on Split Testing here.