Extravision Blogsblog

Recent Posts



Testing, Testing, Testing

September 2, 2011

We all know that testing can help improve our marketing campaigns. However, sometimes it's hard to know where to start and so in this article, we take a look at a few of the top tests you should be running in your email marketing campaigns. 

Before starting your tests you must look at your current situation. What are your open, click-through and conversion rates? Once you have these clearly identified you can start. Make sure that you only change one element of the email at a time; otherwise you won't know which difference in the two emails is changing the way your readers react to them. You should also make sure that one of the emails; the control email; is the way that you would normally write the email, without changing anything.
Each of the following tests can easily be run with split A/B testing, which means that one segment of your list receives one version of the email and another segment receives another version of the email.

Call to action testing

Your call to action is one of the most important parts of your email and has the most to do with your conversion rate. A weak call to action will result in less people clicking through to read your post or look at your advert or buy something. It's a good idea to always split-test your call to action every time you send an email to your list, just to make sure that you're continually improving it.

Time of day testing  

Sending your email at a different time can make the difference between it being opened and read and it sitting at the bottom of a pile of unread emails; only to get deleted . There are no hard and fast rules for the best time to send email - it's all going to depend on your audience and their habits. So test several different times and see which result in the highest open and click through rates.

Subject line testing

Your subject line is as important as your call to action, if not more so, because it has to compel readers to open the email and read it. The best subject lines encourage curiosity, or immediately tell the reader how they'll benefit from opening the email. Subject lines are another element that should be tested as often as possible to make sure that you're continually improving them and that you're getting the highest open rates possible. After all, if people don't open the email, they can't click your links, read your copy, or buy what you're selling.

Desire v Descriptive subject lines

A simple way to test subject lines is to split your database into two segments. To the first segment you will send an email using a "desire" led subject line. To the other you will send the same email with a "descriptive" subject line. It's important that you only change the subject line and keep all the other elements of the campaign consistent to ensure you're only testing the subject line.

The subject lines could take a structure similar to below:
  • "October Newsletter - Some simple sales techniques"
  • "You'll have more sales than ever before"

Notice the difference between the two subject lines. The first sets the scene and describes what the email is about. The second subject line plays on the recipient's desire to grow sales, but doesn't clearly state what the email will cover. When analysing this test we suggest you compare the figures for open rates and click-throughs to see which performs best overall. It is possible that changing the email subject line can increase opens but reduce click-throughs. It's important you analyse as many elements as you can as this will help you form a clear view of what works and what doesn't work for your sector.
Best practice suggests that subject line number one is most effective. However, there has been some research to suggest that "desire" can out-perform "description".

Email length

Many tests have been conducted on length of copy and the affect this has on marketing success. Many of these tests conclude that there is an optimum length for copy. However, they rarely prescribe a specific length for a specific target audience or campaign type. With this in mind, ask yourself if you have ever tested email length to see what works best for your business.

You can test short copy versus long copy in the following way. Segment your email database into two segments. You'll then need to create two email campaigns. The first campaign will be the long version. It will have most of the information on the email and will almost certainly require the recipient to scroll down in order to read it.

The second campaign will need to be much shorter, limiting the scroll to an absolute minimum. You can reduce the amount of copy on the email by using landing pages. For example, rather than writing the whole paragraph, simply summarise it into one or two lines. Then have a "read more" link through to a landing page. Finally, you should send the campaigns to your segments at the same time in order to keep the test fair and consistent.

Analysing these campaigns can be difficult as the benefits of short and long copy are not always easy to identify. We suggest you look for the following key indicators of success:
  • Number of opens per subscriber and as a total of all subscribers
  • Conversions (the number of sales or enquiries generated)
  • Contacts (number of people that reply or contact your business as a result of the email)
Click-through rate is likely to be misleading as the short email will inevitably have more links than the long email. Both long and short copy has its advantages and both styles can work really well.

Image versus text based links

This test is extremely popular with email marketers. Research into this area has identified that there can be a significant difference between the performance of text links and image links. This depends on the environment in which the email is sent and the psychology of the recipient. Even though the environmental and psychological elements are difficult to control this test is still a very useful one to conduct.

Split your database into two segments and make sure the characteristics of each are similar. You then need to create two email campaigns; the first using text links and the second using image links, with these image links being pictures or buttons, depending on what will fit best with your email design. Deliver these two campaigns at the same time, on the same day and then analyse the results. Which campaign has the highest click-through rate? Which produces the most conversions or contacts?

There is another way to structure this test. Design one email. Place several links on the email from images and then the same links from text. Positioning the text and image close together will help you identify which performs best.
Other things that might be worth testing include the pricing of your items, the tone of the copy in the email, or the formatting of the email (lots of graphics vs. very little graphics vs. no graphics). You can also test a certain email with specific segments of your list and see how it converts, compared to the master list.

With each test that you decide to do, try to test something that will give you insights as to what you can use in the future. In time you'll start to develop the ability to predict what your market will respond to, which will lead to higher profits!

If you need further advice about testing your own campaigns, or you would like Extravision to do some testing of your email campaigns, please call 0161 817 2929 or email [email protected] - we're here to help.


Posted by Rebecca McCormick

No comments

Comments are closed.

From the Blog

How to craft the perfect re-engagement email: Topshop example

In an ideal world, everyone on your list would be eagerly awaiting your email. They…

Jenni Malley
June 1, 2017

How To Get Your Perfect Email Template

If you send email campaigns then you probably use a template. Not every email campaign…

Jenni Malley
May 12, 2017

Latest Tweets

Get in touch

MediaCity UK
M50 2AB
Find us on Google maps

Main: +44 (0) 161 817 2929
Support: +44 (0) 161 817 2930
Email: [email protected]


Facebook IconTwitter IconVisit Our BlogVisit Our BlogVisit Our Blog