You don’t have to guess whether your B2B email initiative will be a success.

Just give it a test drive—A/B multivariate split testing, that is.

What is A/B testing? Simply stated, A/B testing is a methodology that splits email recipients into two groups. Each group receives a slightly different version of the email message.

But here’s the trick:

To be statistically meaningful, the split must be done the right way. Every other recipient should receive different content rather than merely splitting the group of recipients in half. In other words, the first recipient gets version A, the second gets version B, the third gets version A, and so on.

But in B2B marketing, split testing is not often used.

Why? Many B2B marketers believe you need to test huge groups—far larger than their email list—to get meaningful results.

Not so. In fact, to be 95% confident that the variable in email version A with a 7.5% response rate is more effective than the variable in email B with a 5% rate, guess the size of the test sample:

• 10,000
• 5,000
• 1,000

Even on B2B-sized email lists—which are typically much smaller than consumer lists—you don’t need a huge sample to be confident. In this case, the answer is 1,000.

To run an email split test, two or more different messages are created and then sent to randomized, equal portions of the larger list.

When the results are in, you’ll know which message is delivering the best response. Then use that message to send to the rest of the list.

Besides better responses, testing also gives a little more insight into your customers. You'll be able to understand factors that might make a difference in future communications. Best of all, you won't need to rely on subjective speculation—you will have hard data in hand.

Anything in your campaign can be tested. For example:

• Subject line: This is the most consequential test subject, simply because it is the first thing email recipients read. Using A/B testing methods, it is easy to create several subject lines and compare open rates.

• Offer: When you're looking for action, you can test various triggers: 25% off may trigger a great response, but what about offering a 10-year extended warranty? Multiple splits make it easy to correlate the most effective offer with various customer segments.

• Call to action: The difference in commitment in clicking on a "Buy Now" link or "Learn More" link is significant. Testing can reveal the most effective time to ask your customers to move from one level to another.

• Scheduling: While there are some rules of thumb on the best times to send email, it pays to test the time of day and part of the week that is best for your customers. Just be sure your email service provider actually delivers the message on time.

• Layout, colors or format: Graphics are not usually as big a factor in B2B direct campaigns as with consumer email. In fact with the prevalence of mobile devices, we've seen significantly higher responses with text than HTML email. Nevertheless, using a single image, a header and other factors can influence readability and clickthroughs.

• Landing page: This can make or break your campaign. Landing page messages should be identical with the email message through the use of Personalized URL (PURL) technology.

But the test must be thought through carefully—if you test more than one variable, you may not necessarily know which one made the difference. It’s all about numbers. That said, don't be afraid to make the tests radically different. If the change is too trivial, it may not make enough of a difference in response, requiring a very large sample size to determine if the change was statistically significant.

So go ahead—put your B2B email plan to the test. You may be surprised at the difference it makes.

Email Testing Tips and Resources

• Email Experience Council
• Free online sample size calculator