Lesson 5

Email Marketing Course (Day #5): Measuring and Testing Your Emails

Well, the email is sent. Time to celebrate and call it a day until next time? Not quite.

I’d argue that the job is, at best, a little over half done.

Now, it’s time to analyze and optimize! There’s always room for improvement, no matter how many times you go through this process. Your next email can always be better than your last.

Defining Meaningful Email Metrics

Like most things in optimization, it all starts with setting smart metrics.

Open rate and click rate are likely the two email metrics you hear of most often. There’s nothing inherently wrong with them, but they can’t serve as your measures of success.

Why? Because they don’t actually tell you anything about your bottom line.

Think about it.

Let’s say you send an email that gets a crazy high open rate. I’m talking 90% of people who receive this email end up opening it. Yet, when you look further down the funnel, you notice only one or two people who opened the email actually purchased something.

Now let’s say you send an email that gets a crazy low open rate. Only 5% of people end up opening it. This time, you see that dozens of people purchased something.

If you’re only tracking open rate, you’re going to think the first email performed the best, despite the fact that it made you less money. The same can be said for click rate. Just because someone clicks through doesn’t mean they’re going to actually purchase.

If you’re not measuring all the way through to revenue, you’re doing your emails a disservice.

Andy Crestodina of Orbit Media suggests another key metric that’s more useful than open and click rate:

“I wish more email marketers would focus on their click-to-open rates. It’s the most underrated email marketing metric. It’s powerful because it isolates other variables and just shows you the effectiveness of your creative and call to action, without much influence from subject line and timing. Very useful!”

TLDR; Open rate tells you how compelling your subject line and preview copy are. Click rate tells you how persuasive your copy is. Neither tell you how valuable or effective the email is.

How to Run Meaningful A/B Tests That Actually Boost Your ROI

Once you’re tracking the right metrics, you can safely A/B test. This is native functionality for most email service providers, which makes A/B testing an email relatively easy compared to, say, a website.

50% of your readers will receive Email A and 50% will receive Email B. Simple! Then, you’ll see which email variation performed best.

Shanelle Mullin of Shopify shared her email A/B testing guidelines with me:

“When I’m A/B testing an email, I always keep the following in mind:

  • Insights are more important than wins. If you have a winner, but don’t know why that variation won, what was the point? You won’t know how to replicate the results and 50% of your readers still saw the variation that didn’t perform as well. If you put the emphasis on gaining insights, every test you run will be valuable, whether your variation wins or loses.
  • Run one test at a time. It’s easy to fall into the trap of testing multiple things at once. If you test too many different things at once, you won’t know what actually caused the win or loss. Test one hypothesis at a time. That doesn’t mean one thing at a time. For example, your hypothesis might be that the messaging is unclear, which results in changes to the subject line, copy and call to action.
  • Declare a winner based on revenue generated. Measure through to the money. Declare a winner based on which variation made you the most money. It sounds obvious, but most people declare a winner based on click rate or, worse, open rate.”

Brian Dean of Backlinko also recommends testing and optimizing your email collection process:

“When I first started my blog, my #1 priority was that it ‘looked nice’. Sure, it was pretty. But it converted HORRIBLY.

So I tried to boost my conversion rate by adding popups, popins, welcome mats and other doodads to my site. These things helped… a little. But it was like adding F1 racing tires to a Toyota Camry. It’s not going to make a big dent.

About a year ago I completely overhauled my site. This time, everything was designed with list building in mind.

My homepage changed from a place to find my latest posts to a page engineered to collect emails. My about page changed from a page to show off to one that now converts at 8%.

And this redesign boosted my conversion rate by 71% (and that improvement has stuck over a year later).”

As you can see, both Brian’s and my own site are focused on growing our list and delivering high-quality content to our readers:

collecting email on homepage



The moral of the story here is that everything can always be improved. Map out your entire email funnel, from email capture through to purchase. You’ll find plenty of optimization and testing points. It’s about so much more than split testing your subject line or call to action copy.

Another worthwhile optimization effort? Segmentation.

How to Master Email Segmentation

Let’s say you have 10,000 people on your email list. 10,000 strangers you know nothing about. Wouldn’t it be more valuable if you knew a little something about them?

Joanna Wiebe of Copy Hackers and Airstory agrees, adding that relevancy is the key:

“The most costly mistake is failing to segment your list and, in doing so, failing to capitalize on the power of really relevant emails. An email is a one-to-one conversation. It’s one of the few ways marketers can have virtually ‘private’ conversations with prospects. When you send unsegmented, low-relevance emails, you are effectively having a crappy one-sided convo with a prospect. And just like at retail / IRL, the prospect will lose interest; unlike at retail, the prospect will walk away midway through your sentence, scoffing and wondering why they ever bothered to consider you.”

It’s difficult to have a relevant 1-on-1 conversation with 10,000 complete strangers.

I know what you’re thinking. “Yeah, Talia, I know segmentation is important.”

The tricky thing is that while most people know segmentation is important, they don’t make it beyond the basics of segmenting by demographic or customer status.

Raphael Paulin-Daigle of SplitBase explains:

“Depending on what you’re selling, some customer segments might see the benefit of your product completely differently than another segment (who could be in the same demographic). Segmenting by the perceived outcome and benefit of what you’re selling should not be overlooked.”

Like Raphael, I encourage you to go beyond the most obvious segments. Here are just a few ideas:

  • Average order size (AOV).
  • Purchase frequency.
  • Loyalty membership.
  • Annual revenue.
  • Job type (freelancer, in-house, agency, etc.)

The list goes on and on, and really depends on the type of business you’re in. Think outside the proverbial box.

Janet Choi of Customer.io breaks it down for you:

“I wish more marketers thought about segmentation and personalization as a way to create a relevant conversation with recipients. Even now, lots of businesses get stuck at demographic or geographic data as standbys for segmentation. The conversation level there is stuck chatting about the weather — hey it’s super cold → now go buy this coat.

Now, this can be super relevant if you’re a clothing ecommerce business. But 1. We’re not all ecommerce weather-affected businesses and 2. There are so many other types of conversations we can have.

Think about what other context you know about your audience and how that fits in with your brand or the overall customer experience you’d like to create. What data or information do you have about previous interactions with your brand? From behavioral context like event data, purchase or browsing history, to where someone is in a buyer’s journey or funnel, to activity level in an app or engagement with previous messages — these are all signals of intent, motivation, mindset, and turns in a conversation that you can respond to and build on.

That’s what personalization comes down to — it’s communicating something relevant, interesting, or valuable to someone. And that’s why great email marketing doesn’t need to mention a first name to be great.”

Another option is to lean on third-party data. Chances are you have a lot of customer and reader data that you’re not employing. Why? As Linda Grant of Vero explains, that data is usually spread across multiple different tools:

“With the ability to collect real-time customer data from our apps, websites and marketing tools – we don’t always seem to spend enough time joining it up, gaining insights and using it to help us design relevant emails campaigns, that customers will actually want.

It’s important to gain a complete and accurate view of all your data and to take time to understand your customer’s behaviour at every touch point in the journey. Each interaction (or lack of) is your customer telling you what they want and don’t want – make sure you’re listening because in a world of information-overload you get very few chances to prove your value. Join up your data and learn from it to keep your customers engaged.”

If you can bring those different tools together with a tool like Hull, for example, you’ll be in a much better position to segment and personalize.

Personalization goes hand-in-hand with segmentation. Once those 10,000 strangers are segmented, you can offer them personalized content and solutions, like Joanna recommends.

How to Master Email Personalization

Don’t just take her word for it. A recent survey found that 78% of consumers are more likely to be a repeat customer if a retailer provides them with personalized offers. The more personalized you can make your emails, the stronger the relationship you’ll build and the higher your conversions will be long-term.

“78% of consumers are more likely to be a repeat customer if a retailer provides them with personalized offers.”

Personalizing your emails can mean:

  • Sending emails out according to time zone or geographic location.
  • Referencing important dates, like a birthday or an anniversary.
  • Sending product recommendations based on previous purchases and searches.
  • Sending tips based on reports previously viewed or features recently used for the first time, for example.

Again, this is very contextual and will depend on your business.

Imagine standing in the middle of Times Square screaming about your brand and the solution you offer. Sure, lots of people are hearing you, but they’re not really listening. That’s what it’s like to send an unsegmented, unpersonalized email.

If you’re talking to 10,000 strangers, you’re not really talking to anyone at all.

Hopefully you’ve taken a lot from this email marketing course.

Have any questions or want some additional content? reach out to me here.

<– Yesterday’s lesson: Sending your emails

Scroll to Top