Steps to Effective Email Testing

Email marketers list testing as one of their top priorities year after year. However, most do no testing at all. At the very least, they rely on unscientific tests and their instincts to determine which subject line, offer, or creative will perform the best.

Testing is not a short-term silver bullet that will solve all your problems. You can't run a few tests and expect your program to hum along profitably for years. What worked well last year probably will not work now.

Marketers who test effectively to achieve significantly higher open rates, click-through rates, and conversion rates are disciplined to run A/B split tests in every email campaign and committed to making changes in their program that make a difference.

In this column, I will explore six steps to effectively test your email marketing campaigns.

Step 1: Ask a Question

What are you hoping to achieve? Determine a specific objective to accomplish for each email campaign. Don't attempt to achieve multiple objectives with one blanket approach.

A series of small steps can be easy to test and analyze. Structure your thinking this way: "I'd like to...":

• Have more people open my messages.
• Have more people click through to the engine.
• Reengage with historically inactive people.
• Have people click on a specific area, topic, or action.
• Drive more conversions.

Step 2: Form a Theory

Use your marketing experience and best-practice knowledge to determine what aspects might make a difference in achieving your objective. "I think...":

• People might be bored with my current subject lines.
• Placement of the specific content might drive more people to action.
• People might not understand this offer/content and therefore will not interact.
• My calls to action need to be stronger.

Step 3: Create the Test

Set up your test, following best practices. Remember, you don't need to prove the obvious.

To optimize opens (one per test):

• From name
• Subject line
• Best time to send

To optimize click-throughs (one per test):

• Creative/layout
• Subject lines
• Offer
• Copy
• From name
• Calls to action

To optimize conversions (one per test):

• Landing pages
• Calls to action
• Creative/layout
• Subject lines
• Offer
• Copy
• From name

Step 4: Split the List

Choose the best list or segment to test, and split it (for that specific test). Which statement below matches your situation?

• I'm confident this list is the most appropriate to prove or disprove my theory.
• My list is large enough to do only an A/B split.
• My list is large enough that I can break it into a larger control and several smaller test segments.
• My list is large enough that I can sample a percentage of my list to test.

Step 5: Measure and Analyze Results

Measure and analyze results to gain insight and prove or disprove your theory. Compile stats (to conversions) accurately. Look beyond the numbers to find meaning.

Even small percentage differences can mean large gains in response rates:

• Opens increased __ percent.
• Click-throughs changed __ percent.
• Conversions changed __ percent.
• Traffic to the website increased __ percent.
• Click-throughs were more focused on specific area, topic, or action.
• Click-throughs were spread out across areas, topics, or actions.
• Conversions increased __ percent.

Step 6: Make Changes

Commit to making at least one change in each campaign. How would you finish this sentence? "I need to...":

• Change my from name.
• Change my subject line.
• Change specific words.
• Change the subject line format.
• Add content.
• Decrease content and simplify.
• Increase clickable areas or links.
• Highlight actionable items more.
• Change copy.
• Modify the layout.

The Last Word

Keep it simple. Initially, test only one area at a time to maintain simplicity and establish a testing regimen. Base your prioritization of success factors on the easiest things to test that will result in the best return.

Establishing a test regimen is the key to unlocking higher open, click, and conversion rates. But don't stop with just one test. Far-reaching conclusions based on one test are not predictive.

Repeat tests where possible over a set time period. Aim for at least three controlled tests before drawing a meaningful conclusion. This will keep you on track for continual growth and improvement in your email marketing program.