In a 2011 case study conducted by AWeber communications, they found that a clear subject line gets 541% more clicks than one that’s clever.
I think clear usually beats clever is because clear subject lines tend to have better consistency into the body of the email, which accounts for higher raw clicks. No tricks, no clickbait, and no question about what the email actually contains.
A few examples of clever subject lines that I’ve seen are “you’re not alone”, “it’s finally here”, and “still doing it the old way?”
Yes, they might get opened out of curiosity but will they be clicked? Hmmmm, not so sure.
For clear subject lines, here are a few that I’ve used: “9 Must-Try Email Subject Lines”, “Your Winning Personalization Strategy”, and “3 Steps to Winning Customer Loyalty”
The clear subject lines might not seem as sexy or luring, but they work exceedingly well.
To prove that clear subject lines work better, I ran a test for our marketing automation checklist ebook.
I tested “Your Marketing Automation Checklist” vs. “Are You Evaluating Marketing Automation?”.
In “Your Marketing Automation Checklist”, there is no question about what the email contains. Sure enough, once you open it, what you thought you’d receive is exactly what you get.
“Are you evaluating marketing automation?” doesn’t say you’ll be receiving a marketing automation checklist, so your subscribers don’t really know what the email is offering. Is it a discount on marketing automation? Is it a super sales-y email from a bad sales rep? It’s hard to tell. But you think maybe the subject line will work well since folks who are evaluating marketing automation surely would need a checklist to do it, right?
Let’s look at the results.
“Your Marketing Automation Checklist”, which was just the name of the ebook, had a 14% higher open rate, an 80% higher CTR, and a 58% higher click/open rate.
Here, we saw a clear email subject line beat out the variation on all metrics.
When we look at this, you might be thinking “well, yeah, this is a no brainer”. But most of us, if not all of us including myself, struggle to write subject lines. How could writing 5-7 words be so hard? We used to write 2,000 word papers in college for breakfast. Some of us can form sentences in our sleep. I know I sleep talk sometimes.
The truth is that when it comes to subject lines, you don’t need to spend hours trying to dazzle your audience with the catchiest subject line of the year. Most often the clear, concise subject lines prove to work better for opens and clicks across the board.
Also remember that your subscribers self identify their own needs. For example, if someone is evaluating marketing automation, they may look at the “are you evaluating marketing automation” subject line and not open it because they thought, “well, yes, I am evaluating but all I really need at this point is a checklist.” By being straightforward with the subject line and keeping consistency throughout the email, we are filling that need.
Tell your subscribers exactly what they’ll get
Front-load the important words
Focus on all metrics, not just open rate
When in doubt, test it out
So why focus on email list hygiene? Well, getting your email delivered is much tougher today than it was five years ago.
In today’s world, avoiding spammy words isn’t enough to hit the inbox. Big internet service providers have followed Gmail’s push towards an engagement model. You may have noticed that Gmail now has tabs: one for primary email, another for social, and a third for promotions. Hitting that primary inbox is increasingly difficult because now Gmail looks at several things:
Continuously opened emails
Unique clicks as well as multiple clicks
Scrolling
Frequency of engagements (does the recipient open emails and engage with them regularly)
So how do we build and maintain a cleaner list for better email deliverability?
So, a sender reputation is the reputation you have as an email sender. Return Path has a tool called Sender Score which rates your IP address based on your email sending practices and gives you a score of 1-100.
You sender reputation is the #1 deliverability reason why you should implement good email list hygiene practices. Due to many recent filtering tactics of ISPs, one thing is very clear; consistently sending emails to a dirty email list will have adverse effects on both inbox placement and your sender reputation.
There are many factors that affect your sender reputation, but the most common ones are subscriber engagement like opens and clicks, positive and negative engagement signals like whitelisting an email addresses or marking an email as spam, hard bounces, if you are listed on a blacklist, spam trap hits, and spam complaints. So being able to manage all of those will help keep your reputation high!
Within your database, it’s likely that you have emails that soft bounce continuously. Soft bounces aren’t usually considered harmful, but those that bounce over and over again can end up becoming something worse in the future.
So to mitigate that risk, you can use marketing automation to run two types of campaigns to manage soft bounces.
First, you’ll want to run a one time batch campaign to clean up any existing emails that have soft bounced repeatedly in the past. In this example, I’m saying that for any email that has soft bounced at least 10 times in the past 60 days, mark that email as invalid, meaning retire it from future campaigns.
Then, you’ll want to run an automated trigger campaign that catches reoccurring soft bounces as they happen in real time. So the campaign listens for emails that have soft bounced more than 6 times in the past 30 days and then marks that email as invalid.
What these campaigns accomplish is two things: (1) you are cleaning up all continuously soft bouncing email addresses from the past up until the present moment. (2) you are automating the clean up of emails that continue to have this issue in the future.
So, this allows you to build a self cleaning system where the more you send emails, the cleaner your list becomes.
High bounce rates lower deliverability and reputation
A good sender reputation helps inboxing rate
Managing soft bounces, role accounts, and unengaged subscribers makes for better list hygiene
FAILURE TO MEASURE EMAIL INBOXING
According to Return Path, only 79% of commercial emails hit the primary inbox. That means 1/5 end up in junk, spam, hard bounce or go undelivered. If you are counting on all 100% of delivered emails to hit the inbox and 1 in 5 do not, that’s a BIG deal!
Also, how many times have you ever looked in your spam folder and clicked a bunch of those emails? My guess is probably not often. When we see that something has been determined spam, we tend to keep our distance.
Did you know that when an email goes to the spam folder or junk folder, it still counts as delivered?
So you might have a 99% email deliverability rate, which seems awesome, but maybe only 50% of those emails actually hit the primary inbox. And if you’re like any business that relies on email marketing to drive revenue, 49% of your emails going to spam or junk is a BIG deal! But how would you know if that’s happening?
Luckily, there are tools that measure inboxing, which is the percentage of delivered emails that actually hit the primary inbox.
Here’s a screenshot from a campaign I ran in Marketo. You’re probably used to seeing metrics just like this. 1,142 emails were sent, 1,124 emails were delivered and your deliverability rate was 98.4%. That’s a pretty great deliverability rate, but as I just mentioned, what percentage of those emails actually hit the inbox and not a spam or junk folder?
This is where an inboxing tool becomes your new best friend!
Here’s a screenshot from Marketo’s integrated deliverability tool called 250ok. As you can see, deliverability is broken down much more granularly than “you got 98.4% email deliverability”. Each row represents a different email provider and the columns show you your inbox rate, your spam folder rate, and the rate of emails that went missing entirely. These domains can be weighted so that if 50% of your email list happens to be Gmail, you get a fair inboxing representation in this tool.
The total inboxing on this campaign was 89.5%, which is very good especially since we know from Return Path that on average only 79% of emails hit the inbox. You can also use this tool to look at inboxing rate broken down by spam filteres.
Introduce bounce management campaigns: this will help you increase your inboxing rates across all ISPs
Measure inboxing across ISPs and work to increase inboxing for specific ISPs where you see more inboxing trouble
Analyse emails for common spam triggers, whether it’s in copy, images, or HTML
NO ENGAGEMENT SEGMENTATION
Getting an ISP to love you is no easy task. Getting all of them to love you is arguably more difficult that getting your celebrity crush to love you. Believe me, I know.
So far, we’ve covered a few points for what makes an ISP love you as a sender:
They love email list hygiene – bounce management campaigns, good opt-in practices, removal of role accounts, etc.
At this point, it probably feels like this relationship with ISPs is a lot of work, huh? And it is! But the #1 thing they love to see is high levels of engagement. That means lots of people opening, clicking, reading, scrolling, and engaging with your emails on a regular basis. When you have high engagement, ISPs will let the majority of your emails hit the primary inbox because the demand for your emails is high!
Okay, so ISPs let emails in and get placed in the primary inbox when your recipients regularly engage with them. So how do you use that to your advantage to get more inboxing?
Let me give you an example. This is just a mock campaign.
Let’s say you have 100,000 emails that are delivered. Within those 100,000 emails, 20k have engaged within the last 90 days. The remaining 80k haven’t shown engagement in greater than 90 days. Let’s look at the first column. If you were to just send to those 20k emails, the open rate would be 18%, the CTR would be 3%, and the unsubscribe rate would be 0%. I’m assuming the unsubscribe rate is 0% because typically when people just engaged, they aren’t likely to unsubscribe. So these are great metrics!
Conversely, if you look at the middle column, if you just sent that 80k of emails who haven’t engaged within 90 days, the open rate would be 3%, the CTR would be 0.2%, and the unsubscribe rate is a little high at 0.31%. These numbers could definitely be better.
When you send all 100k at the same time (the far right column), you have a 6% open rate, a 0.76% CTR, and a .25% unsubscribe rate. Normally, this is probably what you do and these would be the metrics you’d expect based on this example.
Now, what if I told you that if you send the email to the engaged 20k first, wait 30 minutes, and then send the 80k unengaged emails, you’ll get better inboxing rates? Yep, that’s correct. When you send the engaged 20k first, the ISPs will look at you and say “Wow, look at all the engagement on that email. People seem to love this sender and what they have to offer! Let’s boost their reputation and keep it high.” Now when you send the remaining 80k emails, you’ll actually get higher inbox placement on those 80k just because you warmed up your sender reputation 30 minutes prior.
Now look at the bottom row titled “Staggered Sends Inboxing” and compare it to the inboxing row right above it. For the first 20k sent, it doesn’t matter either way. But if you stagger sends, the inboxing increases on your unengaged emails from 55% to 70%! That increases the total inboxing rates for all 100k emails from 63% to 75%! That’s a 19% improvement!
If you stagger your sends by engagement, you’ll see higher deliverability rates and much higher inboxing! This is a really cool trick that not too many people use today, but it is extremely effective!
If you are a high volume sender, consider separating out your email campaigns by engagement
Staggering email campaigns by engagement increases email inboxing
IP address segmentation for engagement can help mitigate business risk
NO RE-ENGAGEMENT STRATEGY
You know that one friend that can sleep anywhere? On a cramped airplane, in the car, on a stool, on the floor, or even standing up? Once they knock out, it seems almost impossible to wake them up.
If I had to bet, it’s likely that there are some email subscribers in your database that are out cold just like this sleeping puppy. And that’s a problem, because you spent time and money getting people to subscribe.
Not only that, these sleepy, unengaged subscribers are bringing down your open rates, click to open rates, email deliverability and sender reputation.
Subscribers become inactive for various reasons, including:
Changed interests
Expectations not set appropriately
Too many emails
Job change
Irrelevant content
Too busy
Changed email address
To wake up sleepy subscribers, we run reactivation campaigns. A reactivation campaign is an email campaign or multiple campaigns specifically targeted towards “sleeping subscribers”, or subscribers that haven’t engaged with your emails in a long period of time.
Why use reactivation campaigns?
Awaken the sleeping subscribers that still want to hear from you
Determine who doesn’t want to hear from you
Clean out your email lists – the older emails become, meaning the longer it’s been since they have engaged with you, the higher the likelihood they could become a hard bounce or a spam trap
And most importantly, if the cost of acquiring a new customer is more than keeping an existing one, you should do everything you can to re-engage inactive customers or subscribers. Reactivation campaigns build upon a brand’s previous investments in acquiring and targeting new customers who already are aware of and have previously engaged with your brand. Re-engaging customers just makes good economic sense.
The BIG idea here is that whatever normal campaigns you are running to these inactive subscribers isn’t working. The language or the offers just isn’t enough to keep them engaged. And if it’s been a long while since their last engagement, they get tired of receiving your emails and mark you as spam or stop engaging all together. Rather than let that happen, you are proactively asking if they still want to receive communications from you.
We run reactivation campaigns here at Marketo and this was our very first test. Our audience was subscribers who had not engaged with our campaigns in 1 year or more. On the left, you’ll see our control, which was for the Definitive Guide to Digital Advertising. There is nothing different about this email in terms of messaging or content for a subscriber that hasn’t engaged in a while. On the right, you’ll see our test.
The subject line for our test email was “First Name, We Really Miss You”. The banner reads “We Really Miss Hearing From You” and we use a sad looking dog to get the emotions going. And to use language that our email subscribers aren’t familiar with, we start with “Here’s the deal, Mike”.
We let them know why we are reaching out. “At some point, you subscribed to Marketo emails. Ever since then, we’ve been emailing you our best ebooks, cheatsheets, definitive guides, and other marketing offers. But then we noticed something disturbing. You haven’t opened or clicked any of these emails in the past year!!!!”
So far, we’ve provided context. This email then goes into saying how we don’t want to clutter your inbox with unwanted email and unless you click the big orange button (which you cannot see here), we won’t know it’s okay to keep sending you emails.
So the takeaway is this: provide context, be straightforward, and ask them to re-engage with your emails. Otherwise, you might just unsubscribe them forever and no one likes missing out.
Now let’s look at the results!
We had a very clear winner here and it was our reactivation email!
We saw a 70% higher open rate, a 325% higher click to open rate, and a 621% higher click through rate. I should also add that we called out the link to our subscription center in this email as well, so about half of the clicks we received were to update email preferences, which is a positive result.
Overall, this email reactivated 238% more subscribers than our control email. That means that we now have a way to engage our sleepy subscribers in a way that is 238% better than our normal marketing efforts!
Use reactivation campaigns to find out who still wants to hear from you…and who doesn’t
Build out a reactivation series of emails, not just one
USING IMAGE-BASED CTA BUTTONS
Alright, so when an email hits your inbox, this is what it normally looks like unless you have images download automatically in your settings. For an email to be counted as opened, images need to be downloaded.
In this example, Live Nation is great as using ALT text so we know what each of these images are before they are downloaded. But what we can’t actually see are the CTA buttons since they are images.
The remedy for this is HTML buttons, which have also been called bulletproof buttons.
So what is an HTML button?
An HTML button is a coded CTA button that looks and feels like an image button but it is just HTML and CSS code. Depending on the browser and device, you can create nearly all of the same effects with code as you can an image-based button.
So what makes this button so special?
Let’s take a look at how Marketo tested into HTML buttons.
Here, we see two emails: the control, which is using a jpeg CTA Button that reads “RSVP NOW” and a Test, which uses an HTML button with the exact same “RSVP NOW” language.
It’s tough to tell which button is which, right? Well that’s because the coded button is made to be the same size, color, look and feel as it’s jpeg counterpart.
Where this button really shines is when the email hits the inbox.
When any email gets to your inbox, you (or your email client) will choose to download images, which counts as an opened email. On the left, you can see that all of the images need to be downloaded and that the “RSVP NOW” CTA Button, the main attraction in your email, is hidden. Only when images are downloaded does your image-based CTA render in all it’s glory.
But on the right, the HTML button renders before the email is opened!!! Why? Because this is code, not an image. Nothing needs to be downloaded at all and your main call-to-action is made clear and apparent before the recipient opens the email.
What’s most interesting about this test is the numbers. Let’s take a peek!
Now right off the bat, you might be thinking “how is it that the HTML button had a higher open rate?” Well, because the coded button rendered before the email was ever opened and it completely changed the user experience. As a result, we saw 5% higher opens!
Additionally, we saw a 15% higher click to open rate and a 20% higher click through rate! What’s even cooler is that these numbers are statistically significant, ensuring that we have full confidence that rolling out HTML buttons to all of our emails will produce a positive impact.
Here’s a few simple, free websites for building HTML buttons:
bestcssbuttongenerator.com
dabuttonfactory.com
css3buttongenerator.com
The takeaway here is simple. Image buttons are not as powerful as HTML buttons.
By this point, it’s very likely you know the importance of mobile responsive emails.
But just in case you need a refresher, here’s some important points for you:
- 65% of consumers start their purchasing path on a mobile device compared to only 25% on a computer and 11% on a tablet.
- Even more impressive, over 50% of consumers open their emails on a mobile phone
So there has never been a more important time to be mobile responsive.
So we took a closer look at the email templates we were using at Marketo. As you can see, things looked pretty good until we got to a tablet size. The issue here was that we weren’t able to accommodate every screen size. So as the email would transition down from desktop to mobile phones, we’d experience a slight break in the banner on the tablet sized version. We had tested this email template to verify that it rendered well across the majority of browsers and devices. We were confident very few people would see the version in the middle.
BUT…
We saw this as an opportunity to explore developing a new template that will help us in the long run. And in a world that’s going mobile, we felt that we had to do this sooner rather than later. So, we made some corrections and crafted a new template that we felt would perform well across all devices, platforms and browsers.
Before I show you the template that we built here, let me first walk you through what components are necessary for a successful mobile responsive template.
Number 1: You need the right media queries. What is a media query? It’s a piece of CSS code (which is code that styles the HTML) that figures out what size screen the email is being viewed on and then displays the code differently to fit that screen.
There are two media queries that we use that cover all the bases, whether it’s a mobile phone (large or small), tablet or desktop.
For tablets, we use “@media only screen and (max-width: 640px)”. This means that when an email is displayed on a screen with a max-width of 640 pixels, it displays differently. And for mobile devices, we change that max-width to 479 pixels. Whether you are an Email Marketing Specialist trying to code a responsive email or a Director of Marketing managing a large team, these lines of code are critical for you to know to achieve mobile success. If you have a dedicated developer, code guru, or marketing maven responsible for your email templates, make sure these lines of code are in there!
Number 2: Mobile devices are smaller. Nobody wants to have to zoom in to read your email. Make the font bigger on mobile devices and your email recipients will love you!
Number 3: Scrolling is fun, but make sure your main CTA appears on the screen when an email is opened. Otherwise, folks have to scroll down to find your call-to-action and you’ll lose out on clicks!
So taking all of that advice, here are the updated templates! We tested them against the ones I showed you two slides ago. They use the media queries we just covered, increase the font size from 16pt to 20pt on mobile phones, and ensure that the CTA remains large above the fold.
Let’s review the results!
Okay, these results were super, SUPER cool because these emails looked so similar from the start.
The new template increased unique clicks by 27%, increased the click-through-rate by 28% and increased the click/open rate by 31%! All of these numbers are statistically significant!
What I want you to take away from this is a couple things:
Mobile responsive is no longer a nice-to-have but an absolute must-have!
It’s not always about the copy or the fun creative images. Think of the template as the vehicle that gets your email to where it needs to be. If that vehicle isn’t working well, your email won’t perform to it’s full potential! Invest some time into optimizing your templates and every single email you send WILL perform better.
What I want you to take away from this is a couple things:
Mobile responsive is no longer a nice-to-have but an absolute must-have!
It’s not always about the copy or the fun creative images. Think of the template as the vehicle that gets your email to where it needs to be. If that vehicle isn’t working well, your email won’t perform to it’s full potential! Invest some time into optimizing your templates and every single email you send WILL perform better.
When running a/b tests on your emails, be sure to normalize your send times. Even a difference of 30 minutes can drastically change the results of a test. Do your best to send emails together and at the optimal send time for your subscribers.
If you’re sample size is too small, you may be calling a winner without actually having a true winner. As a good rule of thumb, I like to make sure I have at least 1,000 observations for every test I run.
So for example, if you’re running a single variable CTA test in your email, you care most about what happens to your click to open rate. Your subject line stays the same between your control and test, so really the only variate is the CTA in the email.
In this case, you want to be sure you have at least 1,000 opened emails for each email to even be in the realm of statistical significance. That means you need to back into your numbers when selecting your sample size. Look at average opens rates for the selected audience and choose the sample size based on being able to predict at least 1,000 opens per email.
Also, run tests 2-3 times to confirm that you do in fact have a true winner. I have found many times that the winner the first time can lose the second time. You need a large enough data set in order to really prove that you can roll out a change to all of your email campaigns without hurting performance.
It’s easy to want to test more than one variable because you think the more you test at once, the bigger the impact you can make and the faster you can make improvements. But the challenge is that you don’t actually know the individual improvements or decline in each variable.
Let’s say we were testing 3 variables at once in an email like the copy, the CTA, and the banner vs. the control email. We run the test with a large enough sample size to see statistical significance and we run the test three times against the control to prove it out.
We find that the test beats the control by 12% on open rate and 36% on click to open rate. We think “great! Now we have a better email!”
Well yes, you do have a better email. But what takeaways do you have that you can confidently say will work across all other campaigns? None. You can’t even say that the open rate is truly 12% better since the CTA and image were different. If the recipient has a preview pane and could see the email before images were downloaded (and you had an HTML button), you open rates could in fact be very different if you were running a single variable subject line test.
My point here is this. Your goal for email testing should be to determine which tests win with statistical significance that you know with certainty you can roll out across all of your email marketing efforts. That’s were you can really move the needle.
If you run multivariate tests and then get a clear winner, you’d have to go back and run single variate tests to really figure out the impact of each variable. Be iterative and your testing practices will be spot on!
Normalize send times
Make sure to have large enough sample sizes for statistical significance (and test more than once)
Isolate your tests to single variables for clean testing
Alright, that’s all I have for you. Before I answer a few questions, I’d like to remind you that there is a brief survey after this webinar. Please take 30 seconds to complete it to let me know how we can make these better for you in the future.
Now on to the questions!
What do you think makes a bigger impact: subject line testing or body copy testing?
How do I convince my boss that if a person doesn’t engage with a reactivation campaign, we should retire their email forever?
Within Marketo, how are you actually able to tell if someone had a last engagement date? And how do you see they types of engagements?
Talk about munchkin code
Campaigns that listen for activity
Campaigns that populate one field for last engagement
My understanding is that not all soft bounces are created equal. Does Marketo have a way of looking at soft bounces by bounce codes and then making decisions based off of those?
Yes, look at emails based on bounce categories. 3, 4, and 9 are the categories we look out for. Some soft bounces mean the email server was busy at that time, which isn’t a big deal.
Do you ever recommend doing a multivariate test for anything?
Yes, to maintain better consistency from subject line to body copy. Some elements need to be paired together in order to work.