The A-Z Guide to Profitable Customer Segmentation

One of the copywriting greats, Robert Collier said, “Always enter the conversation already taking place in the customer’s mind”.

Your emails need to be relevant and contextual, and fit right into the journey your shopper undergoes. 

• What stage of their purchasing journey are they in?
• Where are they coming from?
• What objections do they face? Are they ready to buy?
• Do they trust your brand?

Not all people on your email lists are in the same stage. Let’s take a hipster coffee brand, for example. 

There might be people on your list who love your brand, having consumed several boxes of coffee already. There might also be people who just saw an ad on Instagram and subscribed to your list because you promised them some sweet, sweet discount. 
(There might be several other groups within that spectrum.)

For the second group, you wanna educate them about how your coffee is different, what they can expect, etc. This may be completely irrelevant to the people who’ve been long-time patrons.
Note: You could use the things in this guide for your emails, ads, push notifications, or anything for that matter. But, we’ll focus on email, ‘cause that’s my jam.

Before we get into the approach, mindset, and tactics, let’s define what a segment actually means.

What’s a customer segment?

A segment is a smaller group within the total list of email subscribers, grouped together based on certain parameters such as purchase history, demographics, engagement, and other behaviors and characteristics.

When you send tailored messages to a group of people who all share the same characteristics, you will see better engagement and revenue.

Why should you segment your audience?

The purpose of email marketing is to get more repeat purchases over time. You need a way to send more emails to people without sending every single email to everyone on your list.

Segmentation allows you to scale up your revenue from email marketing by sending more emails over time to the people on your list. You minimize the risk of annoying them (and leading them to potentially unsubscribe or mark you as spam) by being relevant and valuable and tailoring the content to each group. 

The more likely someone's getting email content that resonates with them, the more likely they are to enjoy that email and want to get more emails from you.

But… it’s not just about sending personalized emails

Your segments are also a treasure trove of data that lets you know how your business is performing overall and do a lot of cool things outside of campaigns. 

For example, you could see if your VIP segment is growing and that could be a good indicator of how your retention engine is performing. Your 180-day non-buyers segment–that needs to take a downward plunge.

That’s why the segments in your ESP should include both those who’re subscribed to your lists as well as the ones who aren’t. Unifying this data will help you see the bigger picture.

Here are a few other things you can do with your segment data:

• Branch your automated flows based on the segment. If a customer is in the “non-buyer” segment, their abandoned cart flow should include elements that build trust.
• Show pop-ups on landing pages based on specific segment criteria.
• Export your top customers, create custom audiences and lookalikes, and run ads. That’s what I call “attacking from all sides”!
• Improve deliverability by sending only to your engaged audience until your sender reputation builds up.

Types of Customer Segmentation

Good segmentation starts with good data. The more data points you have about your customers, the better your segmentation will be.

Traditionally, you’d see demographic, geographic, psychographic, etc. But, this is how I like to look at my segments. It puts a lot of things into perspective.

• Customer Attributes: Age, gender, country, city, First purchase date, unresolved support tickets, etc.
• Metric-based: AOV, First Order Value, Total Discounts, etc.
• Behavioral: Purchase history, frequency, site activity, order value, etc.
• Engagement: Email engagement, clicks, etc.
• Zero-party data: Interests, preferences, goals (drink coffee because of the taste or it helps them crank out 1000s of words a day), etc.

Behavioral and engagement data will tell you who your most engaged customers are. If a campaign is likely to be engaged with or not.

Personally, I love zero-party data-based segmentation when it comes to personalized campaigns. Each segment receives content that reflects their interests, making your emails more relevant and more likely to result in revenue. One of the interest-based campaigns we sent for one of our clients, saw up to 54% open rates and 61x ROI.

Here’s a personalized email from Hawthorne based on the data they have about me (or, rather should I say, I gave ‘em ;)).

If you want conversions, sending campaigns to interest-based (or goal based) is the way to go. There’s more nuance to this but we’ll explore all these in future blogs.

What kind of segments should you create?

This is something that’s gonna vary depending on your business, who your audience is, and how, where you operate, and such. 

If you have a smaller list (less than 20k subscribers), it’s better to use basic segments like:

• 30/60/90 day engaged (how you define this will be different if your shoppers have enabled Apple’s MPP)
• Buyers
• Subscribers
• Unengaged subscribers
• VIP segments

However, if you have a much larger list (and resources), you can get a bit more granular with your segmentation. 

Create segments like brand loyalists, high-value buyers, discount shoppers, dormant VIPs, and other engagement tiers to tailor content to those specific cohorts.

The right approach to segmentation

One of the brands I spoke with when starting Forward (unfortunately, we didn’t get the chance to work with them), they sold hair care supplements. 

They had data like which stage of hair loss the shopper was in and what hair type they had, and used that data to segment their audience. They used different sets of creatives and copy in each of their campaigns while sending to these segments. 

Let’s say, Jacob, who was in stage 3 this January, is now getting a lot better. But, the brand has been excluding them from all campaigns that talk about flourishing hair based on their initial goals.

You see what’s wrong here?

It’s a fool’s errand to try and predict customer behavior. We are transient beings; our needs and goals change with time; our beliefs and interests change with time. (Oops, I read too much philosophy this morning).

You’re better off sending emails to a major chunk of your audience. You’ll have much better reach and conversions.

Now, don’t get me wrong. I’m not saying we naively send all our emails to all the people in our lists. 

Us email folks really like to throw around big words like hypersegmentation to sound clever while casually skirting over resources and costs. 

A simple segmentation strategy isn’t sexy and is downright boring. But, our job is to not sound clever; it’s to make money with emails. 

Hypersegmentation: proceed with caution

The more granular your segments are, the better the results are gonna be. Your emails are going to perform better. But…

It only works in a world where we have unlimited money, unlimited time, and unlimited people. We don’t live in that world, do we?

Sure, it’d be great to send a tailored campaign to women between the ages 20 and 30 living in NYC with a red car who have bought 3 packets of Hazelnut coffee in the last 3 months. 

It’d do well, no doubt. Your engagement metrics will shine. Don’t fall for that trap. 

Building the segment, crafting the copy, and creating beautiful designs–won’t be worth the ROI. Using several hours of effort to send a hyper-personalized campaign to 10 people is terrible for your ROI. 

Instead, you wanna find a large group to which your tailored content still makes sense. Like, women who have bought more than 3 times in the last 3 months.

Before creating any segment, I like to ask myself two questions:

1) is it going to be effective?

2) is it going to be efficient?

If the answer is no to any of these, I’m gonna give it a pass.

This is particularly true if you have a small audience and a small team. Think about the opportunity cost; your team might be better off preparing another campaign or running that A/B test you’ve been putting off for a while.

If you have a super-niche brand (like a hipster coffee brand), most of your campaigns should aim to reach the maximum number of people possible (while being relevant and contextual). If your products and audience are diverse, you can experiment with being a bit more granular in your segmentation.

How to get started with segmentation

If you’re using an email and sms marketing platform like Forward, segmenting your audience is easy peasy! You just input a bunch of conditions based on what you wanna build, and voila, your segment is ready to go.

Again, a lot of it boils down to your resources and the size of your list. Start with what kind of segmentation makes the most sense to your brand. What kind of data do you have? What data would you like to collect to improve your segmentation?

For our hipster coffee brand, flavor profiles can be a good place to start. You could also do engaged segments. If you’re in the skincare niche, ask questions like what kind of skin they have and what problems they’re trying to fix (bad acne, dark circles, etc.)

Segmentation is great for driving LTV, repeat purchases, and list engagement. However, weigh the cost of resources against the potential ROI. Once you know that your segments are going to be both effective and efficient, your campaigns will start to get a whole lot better.

‍