Email marketing is my favorite way to make money in eCommerce businesses, especially since you can do so on autopilot. BUT… the only way to truly be successful with email marketing for your eCommerce business is to consistently test to improve your results overall.
Today we’re chatting about my FAVORITE thing… email marketing!
One of my favorite things about email is its ability to make you money on autopilot. You know, all of those email automations that you set up in advance to communicate with your customers and nudge them toward the sale, like your welcome series, cart abandonment, post-purchase, etc.
Don’t have those email automations set up? Check out episode 3 of the podcast here.
The accompanying guide in the resource center will also help to get these set up for your eCommerce business so check that out here.
Automation is amazing. It can help you build a relationship with your customers and make you money while you sleep… but they’re only set it and forget it-ish.
One of the key pillars to email marketing, and my BADASS email marketing framework, is to always be testing.
No matter how much research and strategizing you do when you first set up your email marketing, you will never hit it out of the park the first time. Even if you do, with our lives and environments changing so often—especially right now—your customers’ needs, wants and behaviors are changing as well. What worked really well three months ago might not work as well right now.
To get the most out of your email marketing, you need to consistently improve it. The best way to do that is to test and try new things. This is the case with any marketing activity in your business… hell, even in your operations.
One of the worst things you can do is default to a system, process or strategy simply because that’s the way it’s always been done. We can go down a rabbit hole about the importance of innovation in business (hello Blockbuster) but that’s a conversation for another day.
For today, I really want to focus on improving your email marketing and how you can use testing to accomplish that. Everything listed is relevant for both your automation and your one-off campaigns.
There are 7 main things you can and should test in your emails…
- Segments: Who are you sending your emails to?
- Subject Lines: Your hook to get them to open the email
- Content: What’s inside the email
- Time of Day & Day of Week: When are you sending your emails?
- Frequency: How often you’re sending your emails
- Landing Page: Where are you sending them after they click the email?
- Sender Name: Who is your email being sent from?
Prefer to listen to this episode? Click here
1. Segments: Who are you sending your emails to?
Don’t go too deep into segmentation, especially if your list is on the smaller side. Hyper-segmentation can be overwhelming and actually have a negative effect on your revenue, but if you have some high-level separation in your list, then start using it.
For example, a client I used to work with sells activewear for pregnant and nursing moms. Her very first product was a nursing sports bra, and she has since expanded to leggings. After doing a feature on The View and Good Morning America about the leggings, she collected email subscribers who aren’t necessarily pregnant or nursing moms and so, didn’t need the nursing sports bra. The messaging between the bra and legging segments is different because those nursing sports bras aren’t relevant to everyone on her list.
Another time segmentation can be useful is if you sell men’s, women’s, and children’s clothing. I have no kids, so if a brand is sending me emails about kid’s clothing, it’s not even a little bit relevant to me and I will not open click or buy from those emails.
Don’t go super crazy, but do look for obvious segments in your business. Try testing different messages and product features to these different groups on your list.
You can also segment based on where the customer is at in their journey with your business. I talk about how to do this on Episode 68 of the podcast, so take a listen if you want to explore that more.
2. Subject Lines: Your hook to get them to open the email
This is a pretty obvious one and likely still the test that’s done most often by email marketers, but in my experience it’s not as impactful as some of the other tests we’re going to talk about. Plus, with the rollout of iOS15 looming, we will not have all that open data anyway.
If you’ve never tested subject lines before, then I recommend you do some tests right now so that by the time you lose that open data, you already know what works and what doesn’t.
If you’re not sure what to test, here are a few ideas for you:
- Emojis vs. no emojis
- Witty, funny, or suggestive vs. straight and to the point
- Personalized with their name or the product vs. not personalized
Remember, you can do this in your flows and your campaigns. When I set up a new client’s email flows, I often will add A/B tests so we can figure out what their audience responds to best.
For example, in an abandoned checkout email A/B test, one version might say, “Did You Leave this Behind?” while the other might say, “[Benefit Their Product Offers] Awaits…”
Make sure there’s a good amount of traffic moving through them before you make any final decisions on your flows. If you have pretty high traffic with lots of people moving through them, leave them up for about 90 days before you assess the performance. If you’re newer and you don’t have many people going through, leave them a bit longer. Try to get at least 100 people through, but ideally more.
If you’re going to test in campaigns, I’d rather see you send out the A/B to your entire list with a full 50/50 split vs. sending it to only 20% of your list and then sending the winner to the rest. It’s not an apples to apples comparison and often not as helpful.
In determining the winner, we’re naturally going to be looking at the open rate. And that’s good… that’s what you should be doing. But don’t ignore all your other metrics. For instance, maybe one version got a lower open rate, but it got a higher click rate and generated more revenue. Does that really mean the other one was truly a better subject line?
Hard to say. It’s possible that while that version got a lower open rate it performed better because it was more in line with the messaging inside and the ultimate goal of the email, so it reached a more qualified audience.
I know, I know… I hate when there is a gray area, too. But I swear, I straight up live in the gray. I see all the shades. It’s a blessing and a curse, really.
The thing is, while data is amazing and can tell you a lot about your success, email is equally an art and a science, so you’ll want to use a bit of your intuition along the way.
3. Content: What’s inside the email
The ultimate goal for this step is to increase your click rate and there are a bunch of different things you can test.
Things that influence a click rate include:
- Visually based things like the colors of your buttons
- The layout of your email
- Whether it’s full of images or mostly text based
- Do you have a photo on the top or do you lead with a great headline?
- The calls to action you use: shop now, buy more, see more or learn more?
- Are you telling them to save their cart or complete their purchase?
- Are you including the benefit your product offers in the call to action you use?
- Are you telling them what they’ll get when they click through or are you keeping it vague to pique their interest?
You’ve got a lot of options, friend.
Another thing about content is that so many people will tell you to make sure you only have ONE call to action in your emails. I’m sure I have given that advice as well. And when it comes to best practices, that’s definitely one of them. But… I do know a lot of people that send monthly newsletters with many calls to action that do just fine—so if it works for you, your customer and your business, it’s okay to keep doing it.
I’m working with a client right now who has been sending a weekly newsletter for years with lots of information. When we consulted the first time a couple of years ago, I told her to keep doing it because it was generating so much damn revenue. And at that point, she had already trained her customers to expect it. Plus, it made it pretty easy for her to plan her biz operations, because she knew when that email went out she would get a spike in orders.
Now, a few years later, she is still sending that newsletter email and it’s still generating a ton of revenue. But, her business has grown tremendously, she’s a lot busier now and the weekly newsletter is taking its toll because she takes a long time to put it together.
So, we did a test. She sent 50% of her list the standard long form newsletter. She sent the other 50% a shortened version of that newsletter with about half the information in it.
What do you think the results were?
EXACTLY THE SAME! Literally, open rate, click rate, revenue generated… all exactly the same.
So what does that mean? She doesn’t have to keep creating these super long newsletters. Instead, she can break the information out over 2 emails in the week if she wants, or if she’s short on time, she knows getting out one shorter newsletter will still contribute the same to her business.
Ultimately, when it comes to email, while there are certainly best practices and proven strategies, if your data tells you something different, that’s okay. Do what works for you and your audience.
4. Time of day & day of week testing: When are you sending your emails?
One question I hear most often is when is the best time to send an email? And the truth is, there is no BEST time for everyone. It really depends on your audience and their schedule.
The closest thing I’ve ever seen to a universally best time to send is Sunday evenings between 6 and 9 pm Eastern Time. Which makes sense because most of us are home lounging around on the couch, watching Netflix and scrolling through our phones. If you’ve never tried sending emails on Sunday night, I suggest you try it!
Klaviyo’s Smart Send Time feature
Testing varying days and times can be hard if your list is small.
When I used this feature at my day job, it confirmed what I already knew about my people. Was that because we had already been sending emails at that time for a few years and my customers were trained? Or was it because it really was the right time? I dunno…
Do manual testing
If you don’t have access to
You can also just split your main newsletter list into two random groups. Send the same email to one segment in the morning and the other in the evening and see which one gets better metrics!
This type of test is where you’re really going to need to look at all the metrics together. Just seeing a better open rate or click rate doesn’t mean it’s the winner here.
For instance, you might see a higher open and click rate if you send during lunchtime because people are getting burnt out on their workday and looking for a distraction. But that doesn’t mean they’re ready to whip out their credit card and make a purchase. So, as I mentioned earlier, it’s possible you’ll have one day or time with a lower open rate but higher revenue.
You can use this same strategy to test your day of the week, too. Randomly split your main list and send to one group on Sunday and the other on Tuesday for example.
When testing time and day, cycle through these as you determine a winner and test them against a new hypothesis.
Let’s say you split an email on Sunday, one in the morning and one in the evening. The next week you can split your email against the winning time on Sunday and at the same time a different day of the week, like Thursday.
Again, there are a lot of nuances here as well. You’ve probably heard me tell this story before, but I had two major instances where my tried-and-true Sunday Night emails failed… Super Bowl Sunday and the Series Finale of Game of Thrones.
Make sure you’re thinking about your audience and their typical behavior when you’re hypothesising these tests.
5. Frequency: How often should you send emails?
This is probably one of the TOP questions I get about email marketing. And of course my answer is, it depends.
If you’re trying to figure out how often you should send an email, start with thinking about your product and how your customer uses it. Then the size of your list.
Typically, the larger your list, your product assortment and more frequent your customers use your product, the more emails you can send.
On the flip side, if you’re a one product store and your customer only needs to buy from you once every 4 or 6 months, you can probably send less.
How do you figure it out specifically for your business?
You start testing, of course! And here’s how.
Let’s say you’re currently sending one email per week and you have been doing that for the last few months.
For the next 8 weeks, plan on sending 2 emails per week.
Don’t jump from 1 to 4, that’s going to throw off your subscribers. We’re going to go in small increments here and watch the metrics.
If all of your metrics stay strong, then you can increase again for another 8 weeks.
And then see how far you can push it. But remember, consistency is more important than frequency, so if you get up to sending 3 or 4 emails a week for a few months and then suddenly you drop off the face of the earth, your list may get whiplash ultimately leading to higher churn.
How do you determine whether your list can handle this increase?
You’ll definitely need to look at all your metrics here, including your unsubscribe and spam rates. You’ll want your unsubscribe rate to be less than .3% and your spam rate to be less than .08%. Naturally, you’re also going to be looking at your open and click rates.
Here’s the thing. As you send more emails, those open and click rates are naturally going to go down, but that’s okay as long as your unsub and spam rate stay below those benchmarks and your revenue is going up.
Now, if your open rate gets into single digits, definitely pull back.
If you want to learn more about what benchmarks you should use to assess the health of your eCommerce business and email marketing, check out episode 19 of the podcast where I walk you through all those numbers.
6. Landing page: Where are you sending them after they open the email?
Because yes, you can do all this awesome shit in your emails, but if the page you send them to is shit, or doesn’t align with what was in that email, it will not create a great experience.
If you think about the journey from receiving your email to actually making a purchase, there are a lot of steps.
First, you need to get it into their inbox when they’re likely to see it and have a subject line enticing enough to open.
Then the content inside has to make them want to click through to learn more.
And then the landing page is where you actually get them to convert.
When I say landing page, I don’t mean the typical landing page we associate with sales page used in informational product marketing, which has no header or footer, etc. I just mean the destination you’re sending them to.
Does the landing page make the experience easy for your customer?
Is your landing page any good. Is it doing the selling for you?
Aside from your welcome email, I would say avoid sending them to your home page if you can. It’s just too general and leaves them to have to make more decisions, like where to go and what to click on.
If your email isn’t about anything specific, try sending it to your best sellers category instead of your home page. Remember, you can test this. Create two different content versions. One with a button that links to the home page and the other that links to your bestsellers.
You may also want to test out creating landing pages for big holidays. For Q4, you can create a landing page that links to all of your gift guides, or for Mother’s Day, Valentine’s Day, etc. This should be much easier when you migrate onto Shopify 2.0 (more about those updates in Episode 67) but you can do it now with a page builder like gempages if you need to.
The job of the landing page is to do the selling for you, so create a great experience for your visitors. Is it clear and easy to understand? Do you have the right copy on there? Are your images good and is it relevant to the messaging you shared and the action you want them to take?
7. Sender Name: Who is your email being sent from?
Yup! This CAN make a difference in your results.
Are your emails being sent from a generic email like “info@” or “hello@“, or is it coming from you as the founder or even a customer service person?
Admittedly, I haven’t tested this a ton because my background is in corporate and most of my clients have pretty big businesses, so they don’t want their inboxes clogged up with replies. They’re also hesitant to put their email out there, which I get. But it’s worth testing.
Also, keep in mind that a lot of platforms will let you send from one email but have replies go to another so you don’t have to worry about replies coming to your inbox… but it will make your email visible. Not that a customer couldn’t figure out what it was if they really wanted to, anyway. First name at website url anyone?
In terms of results, I’ve seen it differ across businesses. Some do better when it’s from the founder, others when it’s from a customer service associate. Some even better when it’s just from a generic email because that is what customers expect.
If you don’t want to change the actual email it’s being sent from, you could just change the sender name. So instead of it just being from your brand name, maybe it’s your first name at your brand name.
You may want to only change it for certain emails, like your post purchase thank you. I usually will recommend that you create a text based thank you email after someone makes their first purchase with you. Something that looks like you sat down at your desk and wrote it.
This is a great example of an email that you can make look like it came from you instead of the generic name and email you use for your typical marketing emails.
Test one variable at a time
One last note about testing. Please only test one variable at a time, otherwise you won’t know what is or isn’t working.
So if you want to test subject lines, keep all the other variables we talked about the same between the two emails. Same with everything else we talked about here. Test just one variable at a time.
Start testing now
If you haven’t already been testing these things or you’re not sure what works best with your customer, now is a really great time to start so that you have the data and the insights by the time we hit our peak holiday selling season.
You don’t have to test all the things, just start with the ones that you’re most unsure of and then slowly work on the rest of the tests.