There are a number of weapons in the marketer???s arsenal. One of the cheapest and most effective, if used well, is the email campaign. They can help generate new business and they can turn first time customers into repeat customers. One of the key benefits of an email campaign is that you are already marketing to people who have some interest in your firm (depending on how you have acquired your email list). This means they are already pre-disposed to your firm so as long as you conduct the campaign well you will find it successful.
The key is to run A/B campaigns, that is campaigns where you try to different versions at a time, so that you figure out which techniques and formats work for your email campaigns. This will allow you to hone your marketing strategy so you can improve your conversion rates. Read on to find out how to conduct an A/B email campaign test.
Step One: Work Out What You Are Going To Test
The first part of devising an effective A/B test is to work out what you are going to test. Even if you have a number of different things that you want to test, you need to make sure that you test each individually as it is the only way you can work out whether it works or not. In the scientific lingo, you need to have a single independent variable otherwise you will never know which variable was the one that worked.
There are a number of different things you might want to test, including:
- Call to action (Example: ???Buy Now!??? vs. ???Check Out Our Specials???)
- The specific offer (Example: ???Save 30%??? vs. ???Free shipping on now???)
- Subject line (Example: ???Product A is on Sale??? vs. ???Massive Discounts on Product A???)
- Testimonials to include (or whether to include them at all)
- Message content layout (Example: the use of a single column vs. two column, or different order placement for different aspects)
- Personalization or formality (Example: ???Mr. Jones??? vs. ???John???)
- Body text
- Closing text
All of these will have a different impact on various aspects of the conversion process. For example, the subject line will work on how many people bother to open the email while the call to action will have a impact on how many people click through to your page. This means you need to work out what you want to test first, do you want to try to get people to open the email or are you relatively confident that they will and that it is more important getting them to click through? Also, some aspects are more important than others, the call to action is a lot more vital than the layout of the message so prioritise what you are going to test.
Do You Test The Whole List, Or Just Part?
Ok, the next question is do you test the whole email list or just a part? In general, it pays to do the whole list, but there are some cases where it may be better to just test a part of it. For example, if you have a large email list and the service you are using for your tests charges per email address then you will only want to test a random selection. Another reason you may only want to test a few on the list is if you are doing something really out there and radical so you can limit the potential damage if it does not go well. Remember though that if you are trying something a bit out of the ordinary you will still need to test a decent number of email recipients as too small a sample will not give you an accurate reading, especially of something that will have a narrow but deep appeal.
In general, you will want to ensure that your test sample is in the hundreds if not thousands and that it has been randomly selected. Both of these ensure that the results you get can be scaled up to the entire list, too small a list or one that has been cherry picked will not give you generalisable results. You are trying to get empirical data on your A/B test so you need to approach this scientifically.
Work Out What You Believe Success Looks Like
It is also important that you work out what you think a success is. It is important to do this before you get the results otherwise you may bias the outcome. If you have done previous email campaigns then you should aim to increase the percentage of conversions, if you have never done one before then try to work out what a good percentage would be. With the average open of email generally being 20% and the average click rate being around 2-3%, though this does depend on the industry you are in and, of course, the quality of the campaign.
Tools For Testing
You will find that most email campaign programs have a range of built in tools that you can use for your A/B testing. MailChimp, Active Campaign and Campaign Monitor all have these tools that you can use. If you are using a program that does not have any tools then you can set some up manually. For example, you can split your email list into two separate lists and send one version of your campaign to one list and the other version of your campaign to the other list, meaning that you can just manually compare the results of each campaign. It is important that you are able to easily and comprehensively compare results so make sure that you have the right tools for the job.
Examine the Results
After you have done your A/B campaign you want to examine the results, there are a number of different aspects that you need to focus on including:
- The open rate
- The click-through rate
- The conversion rate once they land on your website
Clearly you will want a good understanding of the first two, but many may wonder if the third is beyond the purview of an email campaign, surely the email has nothing to do with whether they are converted once they get to the site? While in general it is your site rather than the email that got them there the reality is that it is not always so simple. You may think that the email has done its job once it has them click through, but in fact results show that it does. Say you have one campaign that delivers 10% click through and one that delivers 20%, you would expect to get twice as many conversions from the second, but that is not always the case, the better the email aligns with the site the more conversions you will get so monitor the third is important., use some like KISSmetrics to help you track these on-site results.