Email list segmentation splits your subscribers into smaller groups based on what they've done, bought, or told you, so each group gets the emails it's most likely to act on. Instead of one campaign to everyone, you send the offer, reminder, or story that fits where each group is.
Sounds like more work than one big send? It is, a little, but blasting the whole list keeps mailing people who stopped reading months ago, and Google's Email sender guidelines even suggest you consider unsubscribing recipients who don't open or read your messages.
The lift is measurable. When Mailchimp compared about 11,000 segmented campaigns with the same senders' unsegmented ones, segmented sends earned 100.95% more clicks and 9.37% fewer unsubscribes (the data dates from 2017, so read it as direction rather than a forecast).
The segment mix below is where I'd start on any email platform. You'll learn the six segment types, which ones to use most, how to pick your first five, and how to tell if they're paying off.
By the end, you'll have a short list of segments to build this week, so your next campaign reaches more people who buy. Let's start with what a segment actually is.
In this article
- How does email list segmentation work?
- The 6 types of email segmentation
- The segment mix I'd run
- 5 starter segments for any email platform
- How to choose which segments to build
- Common segmentation mistakes
- How do you know segmentation is working?
- Should you get help with segmentation?
- FAQ
- Final Thoughts
How does email list segmentation work?
Email list segmentation works by setting rules, such as "clicked an email in the last 30 days" or "has placed at least one order," and letting your email platform group everyone who matches. In most platforms a segment updates on its own, so people move in and out as their behavior changes.
That's the main difference between a segment, a list, and a tag. A list is where people land when they subscribe, a tag is a label you or an automation attach once, and a segment is a live filter on top of both.
Every segment is built from the same four parts, whatever platform you use:
- A Condition: What the person did or is, like "clicked an email," "placed an order," or "lives in Canada."
- A Timeframe: How recent it has to be, like "in the last 30 days" or "more than 180 days ago."
- AND or OR Logic: AND narrows the group (both must be true), while OR widens it (either one counts).
- An Exclusion: Who should never be in it, like people who already bought the product you're promoting.
Pro tip: Name segments by their rule, not their purpose. "Clicked 30d AND never purchased" still makes sense in six months, while "Spring promo list" doesn't.
The 6 types of email segmentation
Email segmentation comes in six main types, grouped by the data each one uses. You don't need all six on day one, but knowing them helps you spot which data you already have.
| Type | Data it uses | Example segment | Use it |
|---|---|---|---|
| Engagement | Opens, clicks, signup date | Clicked any email in the last 30 days | Most |
| Lifecycle stage | Where they are with you | Subscribed but never purchased | Most |
| Purchase behavior | What, how much, how often they buy | Bought twice or more, last order 60+ days ago | Sometimes |
| Website and email behavior | Pages viewed, links clicked, carts started | Viewed the outdoor category twice this week | Sometimes |
| Stated preferences | Answers they gave you on forms or quizzes | Chose "weekly digest only" at signup | Sometimes |
| Demographic and location | Country, time zone, language, age | Lives in a time zone 8+ hours from yours | Sparingly |
Engagement and lifecycle segments are the base layer, because every subscriber has that data from the first day. Purchase and behavior segments need a store or site connected to your platform, and my guides to behavioral segmentation and RFM analysis go further on both.
Stated preferences are underrated. When someone tells you they only want new arrivals, sending them only new arrivals is the easiest relevance win there is (and it's data you didn't have to guess at).
Demographics are the type people reach for first, and the one that usually tells you the least. Age or gender rarely predicts what someone wants this week as well as what they clicked yesterday does, so I keep demographic segments for logistics like send time, language, and shipping region.
The segment mix I'd run
The mix I'd run puts engagement and lifecycle segments behind nearly every send, uses purchase, behavior, and preference segments for targeted campaigns, and saves demographics for logistics. Here's how that breaks down in a normal week:
- Use Most: An engagement filter on every campaign (usually engaged in the last 60 or 90 days), plus a lifecycle split between customers and people who haven't bought yet. These two decide who hears from you at all.
- Use Sometimes: One purchase, behavior, or preference segment per week when there's an offer that fits, like a refill reminder for past buyers of a product that runs out.
- Use Sparingly: Demographic and location segments, mostly for time zone sends, language, and region-specific shipping or legal notices.
Why lean so hard on engagement? Because it protects the sender reputation every other campaign depends on, and my guide to email sender reputation explains how providers use those signals.
Weak mix: Twelve segments built in one afternoon, with every campaign still sent to the full list "so nobody misses out."
Strong mix: Every campaign goes to engaged 90 days, the weekly promo splits customers from non-buyers, and one targeted send a week goes to a purchase or interest segment.
Pro tip: Add one new segment a month, not twelve at once. You'll learn which ones change results before the list of segments outgrows your ability to write for them.
5 starter segments for any email platform
If you build only five segments, build these. They work in any platform that can filter by email activity and orders, and together they cover the Use Most layer of the mix.
| Segment | Rule (in plain words) | What you send it |
|---|---|---|
| Engaged 30 days | Clicked an email in the last 30 days OR subscribed in the last 15 days | Launches, early access, your best content |
| Engaged 90 days | Opened or clicked in the last 90 days OR subscribed in the last 30 days | Your regular campaign audience |
| Never purchased | Subscribed AND placed an order zero times | First-order offers, reviews, best sellers |
| Customers | Placed an order at least once | New arrivals, refills, loyalty perks |
| Unengaged 180 days | No opens or clicks in 180 days AND subscribed 180+ days ago | A short re-engagement attempt, then suppression |
The last one is an exclusion as much as a segment. Keep it out of regular campaigns, and give it a sunset policy so it gets one honest chance at a win-back email before you stop mailing it.
If you're on Klaviyo, my guide to Klaviyo segmentation has the exact conditions and AND/OR settings for these and seven more. In other platforms, look for segment rules based on campaign activity, signup date, and purchase or ecommerce data.
Pro tip: Base engagement on clicks where you can. Apple Mail Privacy Protection can register opens no person made, which quietly inflates any segment built on opens alone.
How to choose which segments to build
Choose a segment only when you have the data for it, it's big enough to send to, and you'll actually send it something different. If any of those three is missing, the segment just adds a filter to your sending screen.
These four questions decide it for me:
- Do You Have the Data? Purchase segments need order data synced to your platform, and preference segments need a signup form or quiz that asks.
- Is It Big Enough? A segment of 40 people is a spreadsheet, not an audience, so merge small groups until a send to them is worth writing.
- Will It Get Different Content? If the email would be identical, skip the segment and use the engagement filter alone.
- Does It Change a Decision? Good segments change the offer, the timing, or the frequency someone gets.
Weak segment: "Women 25 to 34 in California," sent the same weekly newsletter as everyone else.
Strong segment: "Bought the 30-day supply once, 25 to 35 days ago, no order since," sent a refill reminder timed to when the product runs out.
The strong version wins because it pairs a specific behavior with a specific message. That pairing is where email personalization starts to pay off too.
Common segmentation mistakes
Most segmentation mistakes come from building the segments and then not changing anything about what gets sent. These are the common ones:
- Same Email, Different Segments: Segments exist, but every one of them gets the same subject line and offer. Segmentation only pays when the content changes.
- Overlapping Sends: Someone who's in both "Customers" and "Engaged 30 days" gets two versions of the same promo. Use exclusions so each person lands in one send.
- Opens-Only Engagement: Machine opens keep inactive people looking active. Add clicks or recent signups to your engaged rules.
- Too Many Tiny Segments: Twenty segments of 30 people each are impossible to write for. Merge them by the decision they change.
- No Suppression: Unengaged subscribers still get every campaign, which drags down engagement for the whole list. Clean them up with email list cleaning on a schedule.
Pro tip: Before every campaign, check the estimated audience size. If an engaged-only send suddenly matches your full list size, a rule broke.
How do you know segmentation is working?
You know segmentation is working when a segmented send beats the same email sent to a comparable unsegmented group on clicks and revenue per recipient, without raising unsubscribes or complaints. The cleanest way to see it is a simple holdout test.
Here's the test: split one segment in half at random, send one half the tailored email, and send the other half your standard campaign. Compare the two after a few days, and use the A/B test calculator to check the gap isn't noise.
| Metric | What it tells you | Good sign |
|---|---|---|
| Click rate | Whether the content fit the group | Higher in the tailored half |
| Revenue per recipient | Whether relevance turned into sales | Higher in the tailored half |
| Unsubscribe rate | If you mailed people who didn't want it | Equal or lower |
| Spam complaint rate | Whether the group felt mailed at | Stays under 0.1% (Google's guideline) |
Run one holdout per new segment, then let the winners become your default. For the wider scorecard, my guide to email marketing KPIs covers what to track alongside these.
Weak proof: "Open rate went up after we segmented, so it's working."
Strong proof (example numbers): "The tailored half earned $0.42 per recipient against $0.29 for the standard half, with the same unsubscribe rate."
Should you get help with segmentation?
You can set up the five starter segments yourself in an afternoon, and for most lists that's the right first move. Help makes more sense when your store data isn't syncing cleanly, when you're running lots of segments with no time to write for them, or when you want the segments built straight into flows.
That's the setup my Klaviyo agency team builds for ecommerce brands. If you'd rather see where your current segments are leaking first, request a paid Email DDx audit and fix.
FAQ
What are the main types of email list segmentation?
The six main types are engagement, lifecycle stage, purchase behavior, website and email behavior, stated preferences, and demographic or location. Engagement and lifecycle segments are the ones to use most, because every subscriber has that data.
How many email segments should I have?
Start with five: engaged 30 days, engaged 90 days, never purchased, customers, and unengaged 180 days. Add new ones one at a time, only when each will get different content.
What's the difference between a segment and a list?
A list is where people are stored when they subscribe, and it only changes when someone joins or leaves. A segment is a rule-based group that updates automatically as people's behavior changes.
Is behavioral or demographic segmentation better?
Behavioral segmentation is usually the stronger choice for deciding what to send, because what someone clicked or bought predicts their next action better than their age does. Use demographics for logistics like time zone and language.
Does segmentation help deliverability?
Yes. Sending to people who engage keeps complaints and bounces low, and Google's sender guidelines warn that frequent spam reports lower your domain's reputation and push future mail toward spam.
Final Thoughts
You now know how email list segmentation works, the six segment types, the mix to run, five starter segments for any platform, and how to prove the lift with a holdout test. My recommendation is to put an engagement filter on every campaign and split customers from non-buyers before you build anything fancier.
Your next step is to build the five starter segments today and send your next campaign to engaged 90 days only. Then pick one purchase or preference segment, write it a different email, and run the holdout test on it.
Russ
