• by Aaron Smith , September 12, 2008

The ability to test and track results within emails makes it a compelling marketing channel, providing the ability to continually improve on previous results and bump up that magic acronym: ROI. In the current challenging economic climate, testing isn't a process many marketers can continue to ignore -- it's become a necessity.

 

The good news is: meaningful testing is easier to conduct than you might think, and needn't be disruptive to your regular production process. Effective testing doesn't have to involve complex multivariate data analysis, or the creation of multiple versions of disparate designs. In fact, most times, simpler A/B tests will yield more effective results, and just as importantly won't drive your production team up the walls.

There are several keys to successful testing:

1. Make tests controlled and easy to understand. Simple A/B tests, comparing just one changed element against another, are the easiest to understand and act upon. If you test five different items within the same email, you may need to hire a rocket scientist to figure out the results -- and good luck trying to explain the results to your boss and co-workers.

2. Test frequently and make it a consistent part of the process. Optimally, you should include a simple test in every mail, but at the very least, create a regular schedule for testing (for example, every two weeks or once a month) -- and stick to it!

3. Most importantly, use the results to inform the process. Great tests won't do you any good if you don't act on the results. After sifting through the results, sit down with your team to review the test, and determine how the information will be used going forward.

4. Archive the results. Post the results of your tests in a document, wiki or spreadsheet that everyone on the team has access to. Not only will it help you remember the results of important tests six months after the fact, but it will also be an invaluable training aid when bringing new members onboard.

5. Don't get complacent. We live and work in an ever-changing media world, where today's best practices become tomorrow's pitfalls overnight. Don't be hesitant to run the same tests every few months. Effective subject lines, for example, are a constantly moving target, and what works well today may not be very effective in three months.

With the above key elements in mind, here are a few simple ideas for tests you can run within your own email program:

Graphical versus HTML text: A frequent topic of debate -- best practices say HTML text will generally outperform graphical text. But unless you do a test, you may have a hard time convincing your design team of the efficacy of HTML text.

Text links versus button calls-to-action: We all know buttons perform best for calls-to-action, but some designers may be hesitant to place a button in the middle of their lovely design without some good numbers to help back up the decision.

Subject lines: This is one of the most important tools to encourage email opens, so testing of subject lines should be an ongoing and regular part of your email program. Need I say more?

Best time of day/best day of week to send: As my fellow Insider authors frequently write, the best time of day and best time of week to send messages varies by industry and even by company. There is no magic bullet -- or more to the point, the magic bullet is for you to find out, by conducting a series of tests.

These are just a few ideas -- the possibilities are truly limitless. What are some simple, successful tests you've run in the past? Don't be shy; let us know!

Good luck and happy testing!