Why Test Email Campaigns

In the scope of email marketing testing is a great opportunity to improve the impact of the email campaigns, and ultimately, increase revenue.

Testing is interrelated with segmentation. If you segment your list into smaller groups using some criteria (by the product name, or list name, or frequency, etc.), testing is the only way to know what each group is looking for. Once you identify what they want, you can send them more relevant information, keeping your statistics high and complains low.

In addition, during testing you can reveal and solve specific problems in your emails and ensure they look perfectly for your recipients.

But many email marketers still do not test their email campaigns. It’s not because they do not care about the success of their mailings. Most common reasons for not testing are:

• they don’t know how to test;
• system/software doesn’t provide them with testing capabilities;
• campaign timeline is too short and they don’t have time to work up different testing versions, copies, or designs;
• and lastly …they don’t see any value in testing.

On my opinion, any good email marketer would purposefully refuse to test his email campaigns. As you see two main reasons why email marketers don’t run tests are a lack of knowledge how to do this and lack of email system capabilities. This means they realize the importance of testing in email marketing and this is a good sign already.

What to Test in Email Marketing

Those marketers who are testing their email campaigns are testing a variety of elements such as subject lines, calls to action, email designs, offers, frequency, and sending day/time.

You know in email marketing what is more important is getting your readers to open the email and actually respond to the call to action. Since the Subject line is one of the factors that determine whether the recipient will open the email or not, it’s no surprise that the majority of email marketers are testing their Subject lines. The second place in testing take calls to action. Then go the email copy design, offers, date and time of the email campaign, and finally, email frequency.

During the test some email marketers discover that sending mid-day is the best and the start of the working day is the second best time. As to days of the week, Tuesday and Wednesday seem to be the most successful days for emailing.

However, this is not true for all email marketers. If you offer some goods or services for leisure-time, recreation or sport, you’ll do the right thing if you send your emails during the weekend when most people have rest and will probably be interested in what you are offering.

But once again, the best time/day of sending is individual for each email marketer. What works for one may not work for others. That’s why we are testing our email campaigns – primarily to find out what is the best for us and for our subscribers. The main goal of testing is to optimize our email campaigns to get a warm response from our subscribers, and finally, increase revenue.

How to Test Email Campaigns: Single Variable Testing vs Multivariate Testing

There are two ways to test emails: testing one variable at a time and testing multiple variables at a time.

A single variable test is commonly referred as A/B split testing. A/B testing implies sending several copies of an email differing by a single variable only to a control group to see which copy gets a higher response from the recipients. The winner email copy is sent to the rest of the list then.

The advantages of A/B testing are:

• Quick and easy. You can set up and run the A/B test in a short period of time.
• Clear and straight. Since you are testing only one variable at a time, you can clearly understand which version drives the highest response.

Multivariate testing allows you test several variables at once and can be also useful because:

• Testing many variables at once you can think out-of-the-box, be more creative and try changes you normally wouldn’t try. This can lead to grand discoveries.
• You learn more in less time. Analyzing the effect of several changes at a time you discover more about your subscribers within the same timeframe.
• You can track interlacing effects of variables on one another. For example, how different Subject lines can lead to a higher click-through rate with different calls to action, or how subscribers respond to different offers in the emails sent at different times of the day.

How big should your control group be? The more recipients, the better. To get statistically viable results, a test group should contain at least a few hundred subscribers, but more is always better.

How to Do A/B Split Test Using G-Lock Analytics and G-Lock EasyMail

If you are serious about testing your email marketing campaigns, you can use our G-Lock Analytics tracking system and our email sending program G-Lock EasyMail to conduct the A/B split tests. G-Lock EasyMail is the first in the world desktop email software that supports the A/B split testing capability.

To do an A/B test, you create several versions of the email (2 or 3 versions are enough) differing by one element, for example, by the Subject line, or offer, or call to action, etc. You add the email versions to the Split Tester module in G-Lock EasyMail. You also define to how many % of your list test versions will be sent (usually 10-20% are enough). You can also define how the winner email version will be determined – by the number of opens, clicks, or you will choose the winner version yourself, and set the time interval during which the statistics to identify the winner email will be collected. 

Now you hit “Send”. After all test versions are sent, G-Lock Analytics starts collecting tracking metrics for each version. You can watch the statistics in the Split Tester module in the real time.

When the winner email is determined, G-Lock EasyMail sends it to the rest of your list. Here you can learn more about how to do an A/B split test using G-Lock EasyMail and G-Lock Analytics.

If you don’t have G-Lock EasyMail yet, download a free version now and start testing your campaigns right away. If you don't have an account on G-Lock Analytics, you can create it here.

If either version wins, you can come up with the ideas why it was more appealing for your recipients and use your theories in the future. But again, do not rely on the results of one test only. What worked today may not work tomorrow. Remember that optimization is not a one-time process. Run tests as often as possible, ideally before each email campaign.

Now I’d love to hear your comments and thoughts. What did you try? What worked best? What failed? Leave me a comment below and let’s discuss.
 

Tags: email list segmentation, email tracking metrics, split test statistics, test email campaigns, track email marketing campaigns

Related posts

• Tutorials

Filed under: Email Tracking