Overview
WSO2 sends several weekly and monthly newsletters for various products. The main goal of a newsletter is to build brand awareness and the biggest challenge we encounter is increasing the click through rate (CTR). Most of the time I design templates based on the competitor analysis, [1] current trends, and the previous newsletters engagement statistics. Whilst the main design remains consistent throughout the newsletters, I make constant design improvements [2] based on user behavior and AB testings.
You may be interested to learn that I’ve written 2 blogs based on my research findings about newsletter design. Read them here:
[1] "How I Create Dark Mode Friendly Mailers" and
Design process
This is the design process for newsletters. I take the lead from steps 3 to 6 and 8, making the design deliverables while communicating internally with the marketing content team and externally with developers (our main target audience).
Design layout
My job begins after the respective internal teams finalize the newsletter content. Since 2018 until today, we’ve changed our content and design strategy several times in order to increase the CTR, reader engagement, and address our findings. With every change, I create a new theme to achieve our collective goals.
​
Background research is my starting point, it is only after this that I start to create the layout. I look at competitors’ newsletters, design trends, and analyze past newsletter performance data, paying particular attention to CTR reports. I then send an initial design of the newsletter which I later amend according to the feedback provided by the marketing content and developer teams.
May 2020 - Present
Oct 2019 - March 2020
May 2019 - Sep 2020
Oct 2018 - April 2019
Oct 2018 - April 2019
Developments + Improvements
I use HTML and CSS to develop newsletter templates. When making design changes, I look at several aspects - code improvements, dark mode, layout changes, the use of emojis in subject lines, and what types of CTAs work the best.
Code improvements
One of my team members once attended a Litmus email marketing conference and got an improved version of the code from a conference participant that she shared with me. I used this code to make improvements to the next newsletter that I designed (this was in October 2018).
Observations
Initially, there wasn’t a noticeable change in the email open rates.
Dark mode improvements
Dark mode is a new trend in the tech industry, popularized due to its benefits (lower energy consumption and better readability). Many emails support dark mode nowadays. A common misconception is that this is something that can be done easily.
​
I've changed the logo, images, icons, and colors (text and background), and used custom code (use “prefers-color-scheme: dark” media query) as dark mode improvements. Read my blog for more information.
Observations
Although I haven’t seen a huge improvement in our CTR statistics, I’m sure that these improvements will help to provide a better user experience for dark mode users. I expect improvements once Gmail starts using dark mode and am keeping a lookout for this!
.png)
Layout improvements
I believe small design changes have a great impact on user behavior.
​
#1 Use GIFs rather than still images
I’ve experimented with this many times. First, I sent one email with a GIF (animated) image, a second email with an illustration, and a third email without an image (text-heavy HTML). I also pay particular attention to the first link on the newsletter, as this is often the link that has the highest clicks.
Observations
We improved clicks from 122 to 293 when GIFs are used. I can't say this is only because of the GIFs, it could be the topic areas as well. Our research found that GIFs had a good impact on email performance statistics.
#2 Design changes for specific sections
One newsletter layout had a section where we used to include 5 links to different resources on average. However, we discovered that not many people were clicking on these links. We couldn’t reduce the number of resources that we included in this section, therefore I changed the design layout of this section so that it is more appealing to the eye.
Observations
We improved clicks from 0-28 to 106. This was a combination of the improved layout and helpful reader information included i.e. type of content and estimated reading times.
January 2020
.png)
Oct 2019 - March 2020
Email Testing
We test our emails in two ways. First, the newsletter team and I test the layout, link functionality, and responsiveness via our email automation platform. Then we prepare mailers for A/B testing. At the beginning of the process, we decide what A/B testing we will do for a particular email and I’m mindful of this when I design the templates. ​
We have run many A/B testing so far. You can read more about my experiences and findings in this blog.
#1 Personalization increases open rates and CTRs
I analyzed 4 newsletters to test personalized subject lines. First, our team sent out 2 email campaigns, the first email was sent with the product name (as per our usual practice) and another one with the recipient’s name.
Observations
The email with the recipient’s name worked better in this instance.
#2 Shorter subject lines increase email open rates
First impressions matter. Include the sender name, newsletter subject, and pre-header in the subject line itself to give users an idea of what they can expect from the newsletter.
​
Lengthy subject lines are not fully displayed on mobile devices. If most of your target audience reads their newsletters on their mobile devices, a long subject line will have a negative impact on your email open rates.
Observations
Email subject lines with 6 to 10 words increase your open rate by 21%.
#3 Emojis on subject lines and preheaders
Emojis grab more attention, especially these days where we receive many emails in a day. Some studies have found that subject lines or pre-headers with emojis have a positive impact on the email open rate.
​
We did some A/B testing for emojis by sending the same email, with and without an emoji. Yet the subject line was different on these 2 emails.
Observations
Although the email with the subject line only worked well in our A/B tests, I’m not entirely convinced. I believe it’s best if A/B tests for emojis are done with the same subject line.
#4 Button vs text links
We’ve done several A/B tests to find out if button or text links work better. I thought button links would work better, but the data showed the opposite.
​
Observations
Our A/B tests showed us that text links work better. However, we did send out 2 emails with 2 different CTAs (we used a button with the words “Register Now” and a text link with the words “Save My Seat.”) I believe the testing would have worked better if we had used the same phrases for the CTA links.
#5 Using the same link on different elements to increase the CTR
I added the same link to different elements within the newsletter - the GIF image, main title, text link, and the CTA - taking into account user behavior. We all click on different sections of a newsletter, some of us click on images while others click on text or the main CTA. We used embedded UTM links to track these clicks.
Observations
We saw a significant increase in the CTR by 1.5%. This is now a standard practice across all newsletters. When we analyzed the UTM data, we found that not everyone clicked on the main CTA.
Analysis
This is one of my favorite parts because we learn so much when we analyze user behavior statistics and this helps me to improve the design for the next newsletter. I’ve listed some of my observations below.
#1 The best day of the week to send emails
I tested which day of the week would have better email performance rates with 2 different audiences because user behavior varies across groups and individuals.
​
Observations
As you can see from the results below, there’s no one particular day of the week that marketing emails perform the best. Different results vary according to the day of the week. Plus, it’s important to keep in mind that your audience will change their behavior. So I believe constant A/B testing will help you keep track of any changes.
.png)
Open Rate - Wednesday and Monday
CTR - Friday and Wednesday
CTOR - Thursday and Friday
Open Rate - Wednesday and Monday
CTR - Friday and Wednesday
CTOR - Thursday and Friday
#2 Titles and first links featuring the 5 Ws receive the highest clicks
I analyzed 3 mailer campaigns that were sent to the same target audience. All 3 campaigns had 5 main segments with a CTA each. I also analyzed other email campaigns with other 6 segments. In all of these newsletters, I found that the initial link receives the most number of clicks.
​
Observations
A sample of our results are shown below. Newsletters containing the 5Ws and 1H (why, who, which, what, where, when, and how) in their subject lines and titles have more clicks compared to others. However, this does depend on the audience and interest area. As I mentioned earlier, different target groups have different behaviors. We must identify these by analyzing email user behavior frequently.
.png)
Click-through report of three mailer campaigns
.png)
.png)
Click-through rate analysis
Reflections
Designers and marketers spend a lot of time wondering if their email marketing campaigns work according to their intentions.The best way for us to find out whether or not our campaigns are working as intended is by collecting actual customer engagement data. Before starting any email marketing campaign, clearly define your target audience and marketing goals.
​
You can gain a good understanding of what you’re doing right and what should be improved by tracking email open rates, click to open ratio (CTOR), CTRs, and user interaction. Design and content can always be improved to achieve better user engagement. Working with real data helps you to build a complete story for your campaign performance at every stage.