Email Segmentation: How to Segment Your Email List
Email segmentation means dividing your email list into smaller groups that have something in common, then sending each group what is relevant to it. The thing in common might be what people bought, what they clicked or what they asked to hear about. You will also see it called list segmentation or audience segmentation.
The alternative is sending every campaign to everyone. That works for a while on a new list, then engagement fades and complaints rise.
Two things have changed how segmentation is done. Opens are no longer a dependable signal, so engagement segments now lean on clicks and purchases. And mailbox providers enforce complaint limits, so irrelevant email carries a real cost. This guide covers what to segment by, which segments to build first and how to keep them accurate.
What is email segmentation?
Email segmentation is the practice of grouping contacts by shared data and sending each group its own message. The group is the segment. Segmented emails are the campaigns you send to it.
It helps to separate a segment from a list. A list is a fixed set of contacts that you add to and remove from yourself. A segment is a saved filter, such as "clicked an email in the last 90 days". Nobody maintains it by hand. A contact who clicks today joins it, and a contact whose last click was 91 days ago drops out.
Picture a list of 10,000 subscribers. Some signed up yesterday, some have bought from you for three years, and some have not clicked anything in six months. The new subscriber needs an introduction. The long-standing customer deserves first access to something new. The silent subscriber needs a different question altogether: do you still want these emails? One campaign cannot do all three jobs.
Why email segmentation matters
Relevance is the obvious reason. An email about something you care about gets read. One that has nothing to do with you gets deleted or reported.
The less obvious reason is deliverability, which is whether your email reaches the inbox at all. Gmail, Yahoo and Microsoft expect bulk senders to keep spam complaints under 0.3%, and Google's own target is 0.1%. On a 10,000-address send, 0.3% is 30 people pressing "report spam". Our guide to the sender requirements has the detail.
Segmentation is the practical defence, because the people most likely to complain are the ones receiving email that was never meant for them. Complaints also wear down your sender reputation, and that affects every email you send afterwards.
You will come across precise claims about how much segmentation lifts opens and clicks. We do not quote one. In the accounts we work with at MailGraf, campaigns sent to a few relevant segments usually do better than single-list sends, but the size of the gap differs from one account to the next. The comparison that matters is with your own previous campaigns.
Six ways to segment an email list, with examples
What you segment by depends on your business and on the data you hold. These six types cover most email list segmentation.
| Segmentation type | Example data | Example campaign |
|---|---|---|
| Demographic | Country, city, job title, company size | An event invitation for contacts in Manchester |
| Engagement | Clicks, recent opens, engagement score | First access for people who clicked recently |
| Purchase history | Last order date, number of orders, total spent | A care guide for people who bought a particular product |
| Lifecycle stage | New subscriber, active customer, lapsed customer | A welcome series for new subscribers |
| Stated preference | Topics chosen in a preference centre | Event news only for people who asked for it |
| Sign-up source | Website form, shop till, event, landing page | A welcome email that mentions where you met |
Demographic data is the easiest to collect and the weakest guide to what someone wants. Knowing a subscriber is 32 tells you little. Knowing they clicked your last three emails and bought twice this year tells you what to send next.
Stated preferences deserve more attention than they usually get. A topic someone ticked themselves is accurate, and it involves no guesswork on your part. An email preference centre is where subscribers set those answers and change them later.
Whichever types you choose, segmentation is only as good as the data behind it. If your sign-up form collects nothing but an email address, your options are thin. Decide what you want to segment by first, then make sure you are collecting it.
Engagement segments when opens are unreliable
For years the standard advice in email segmentation was to split a list into openers and non-openers. That split is less trustworthy now. When an Apple Mail user turns on Mail Privacy Protection, the sender cannot tell whether the email was opened (Apple). An email can be recorded as opened when nobody read it. Our A/B testing guide explains why open rates need a pinch of salt.
A click is different. It takes a deliberate action, which makes it the stronger sign of interest, and a purchase is stronger still. Build engagement segments on clicks first and treat opens as supporting evidence.
| Segment | Rule | What to send |
|---|---|---|
| Engaged | Clicked in the last 90 days | Your main campaigns, plus first access to anything new |
| Opening but not clicking | Opened in the last 90 days, never clicked | Shorter emails with one clear button |
| Going quiet | Clicked before, but not for three months | A re-engagement series |
| Never engaged | Sent at least one email, never opened | One last check, then stop sending |
The 90-day window suits a sender who emails weekly or fortnightly. If you send a monthly newsletter, use a longer window; six months is reasonable.
A three-email re-engagement series
The going-quiet segment is worth one deliberate attempt before you give up on it.
- Remind. Send your most useful recent content or offer, with a subject line that says plainly what is inside.
- Ask. A week later, ask what they would prefer: fewer emails, different topics or a break. Link to your preference centre.
- Confirm. A week after that, say this is the last email unless they click to stay.
People who do not respond come off your regular sends. Keeping them does harm, because silent addresses pull down deliverability and hide how the rest of your list is doing. Our email automation guide covers building the series as a flow that runs on its own.
Customer segmentation with purchase data
If you sell online, purchase history gives you your most valuable segments. The usual method is RFM, which stands for recency, frequency and monetary value: how recently someone bought, how often, and how much they have spent. RFM analysis sounds technical. For email segmentation it comes down to four groups.
| Customer segment | Rule of thumb | What to send |
|---|---|---|
| New customers | One order, placed in the last 30 days | A thank-you and a reason to come back |
| Repeat customers | Two or more orders | Early access and loyalty rewards |
| Top spenders | Total spent above a figure you choose | A personal thank-you, previews, a request for feedback |
| Lapsed customers | Ordered before, but not in the last 180 days | A win-back offer |
Set up the new-customer segment first. A first-time buyer has not formed a habit yet, and a well-timed second email does more here than anywhere else. A clothing retailer might write: "Your jacket is on its way. Here are three scarves our customers wear with it."
Purchase data also drives basket reminders. Those are triggered by an event, an abandoned basket, not by a segment, so they belong with your automated flows.
How to start: four steps
Start email segmentation with three segments, measure, then add more. Twenty segments on day one is the usual reason these projects stall.
- Pick three core segments. New subscribers (added within the past month), engaged contacts (a click in the past three months) and quiet contacts (no click in that time). If you sell online, add repeat customers as a fourth.
- Collect the data. Add one or two questions to your sign-up form, give subscribers a preference centre and connect your shop so orders arrive automatically. Collect consent properly as well; double opt-in gives you a confirmed address and a clear record.
- Build live segments, not fixed lists. A saved filter updates itself. An exported list is out of date within a week.
- Send, measure and adjust. Compare each segment's clicks and orders with your earlier whole-list sends. Our guide to email marketing metrics explains which numbers to trust. Add a segment only when you have something different to say to it.
New subscribers are the easiest place to see a result, because a welcome email series reaches people when their interest is highest.
Segmentation vs personalisation
Email segmentation chooses who gets an email. Personalisation changes what is inside it for each person. "Send the spring range to customers who bought in the last year" is segmentation. "Hi Sarah, the bag that goes with your dress has arrived" is personalisation.
The two meet inside the email itself. With dynamic content, one newsletter shows a row to members and a different row to everyone else, and the rule on each row is a segment. Our guide to dynamic email content shows how to set that up and test it.
How email segmentation works in MailGraf
In MailGraf a segment is a saved set of rules, and its membership updates on its own as contact details and activity change.
To build one, open Segments, click New segment and add filters. The fields are grouped: Contact covers details such as country, source, permission and dates. The other groups are Custom fields, Lists and tags, Engagement and Preferences. If your shop is connected, order data is available too, so a rule can read "Ordered in the last 90 days". Join filters with And or Or, and watch Live preview to see how many contacts match as you edit.
If you would rather not start from an empty box, Ready segments offers prepared rules such as "Added in the last 30 days", "Double opt-in contacts" and "Missing first name". Each card shows its rule before you use it.
Two details help with accuracy. Last device, Last email client and Last location are taken only from genuine opens and clicks, so privacy proxies such as Apple Mail's do not overwrite them. And the engagement score, which runs from 0 to 100, gives recent activity more weight than old activity.
A saved segment does four jobs. It can be the audience of a campaign, the rule on a row inside an email, the trigger for an automation, or the starting point for a fixed list. After a send, the report can also prepare a follow-up for the people who clicked or the people who did not. And if you have connected an AI assistant to your account, you can ask it in plain English to save a segment, for example "contacts who clicked in the last 30 days".
The help article Create a segment walks through the builder, and the segments page shows what else they can do.
Common email segmentation mistakes
- Too many segments. Every segment needs its own message. Twenty segments and one newsletter is extra admin with nothing to show for it.
- Segments too small to learn from. Forty people cannot tell you whether a subject line worked. Keep segments large enough for a result to mean something.
- Relying on opens alone. Opens overstate engagement. Check clicks and orders before you call a segment active.
- Fixed lists that go stale. Someone who was engaged in spring may be silent by autumn. Use saved filters that update themselves.
- Guessing at sensitive things. In the UK, the ICO (Information Commissioner's Office, the data protection regulator) says that inferring health, religion or similar special category data to target marketing needs explicit consent. Segment on what people told you. Our UK GDPR guide covers the basics.
- Segmenting without data. If you collect only an email address, there is nothing to segment by. Data collection comes first.
FAQ
What is email segmentation in simple terms?
Email segmentation is dividing your email list into smaller groups that share something, such as what they bought or clicked, and sending each group a message that fits. It replaces one email for everyone with a few emails that are each relevant to the people receiving them.
Should I segment my email list or send the same message to everyone?
Segment it, even if you start with only two or three groups. A single message to the whole list is fine for news that concerns everyone, such as a change of opening hours. For offers and regular newsletters, people who receive email that is not relevant to them stop clicking and start complaining.
What are the four main types of email segmentation?
The four most common types are demographic (country, job title), engagement (clicks and recent opens), purchase history (orders and spend) and lifecycle stage (new, active, lapsed). Stated preferences and sign-up source are two more that are easy to add. Most senders combine two or three.
How many segments should I start with?
Three is enough: new subscribers, engaged contacts and quiet contacts. Add a fourth for repeat customers if you sell online. Add more only when you have a different message for the new group.
How do I segment people who open but do not click?
Build a segment with two rules: opened in the last 90 days, and never clicked. Treat it with some caution, because privacy features can record an open that never happened. Send this group shorter emails with one clear button, and move anyone who stays silent into a three-email re-engagement series.
What is the difference between a segment and a list?
A list is a fixed set of contacts that you add to and remove from yourself. A segment is a saved filter whose membership updates automatically when contact details or activity change.
What is the 80/20 rule in email marketing?
It is a rule of thumb, borrowed from the Pareto principle, that a small share of your subscribers accounts for most of your email revenue. The exact split varies. Email segmentation is how you find that group, through order count and total spent, and treat it differently.
Originally published: Apr 9, 2026


