Email Marketing

Email Segmentation for eCommerce: The Segments Worth Building First

The email segments an eCommerce store should build first, how to define each one, how many you really need and how to prove segmentation pays off.

TL;DR
  • Most stores need six or seven segments, not sixty. Start with an engaged sending audience, a buyers vs. non-buyers split, one-time buyers, repeat customers, lapsing customers and a suppression segment.
  • A segment earns its place only if it changes what you send, when you send it or whether you send at all. If two segments get the same email, merge them.
  • Judge segmentation on revenue per recipient and margin per send, not open rate. Keep a small holdout when you test a new segment so you can see what it actually adds.

What is email segmentation, and why does it matter for profit?

Audience segmentation in email means splitting your list into groups defined by behavior or attributes, then sending each group something different. In Klaviyo and most modern ESPs, segments are dynamic: people move in and out automatically as their data changes.

Segmentation pays in two ways.

  • Revenue per send goes up. A repeat customer and a subscriber who has never bought need different messages. The first needs a reason to come back; the second needs a reason to trust you.
  • Deliverability is protected. Mailbox providers watch how recipients react. Sending every campaign to people who never engage drags down inbox placement for everyone, including your best customers. This is why list hygiene and segmentation are the same job seen from two sides.

There is also a cost side. Discounts sent to people who would have bought at full price are margin you gave away. Segmentation lets you hold offers back from the customers who don’t need them.

Segments are different from flows. A flow is an automated sequence triggered by an event, such as an abandoned cart. A segment is a group you can send campaigns to, use as a flow filter or sync to ad platforms. This article is about the groups, not the flows.

Which email segments should an eCommerce store build first?

Build these in order. Each one changes a decision, and together they cover most of the value. The conditions below are starting points to adjust to your own buying cycle.

Segment Example definition (adjust to your cycle) What it changes
Engaged sending audience Clicked, visited the site or ordered in the last 90 days, or joined in the last 30 Who receives regular campaigns
Never purchased Subscribed, zero orders Message: trust, proof, first-order reasons
One-time buyers Exactly one order Message: second-order push, product education
Repeat customers Two or more orders Message: new products, early access, less discounting
Lapsing customers Ordered before, no order in roughly 1.5–2× your typical gap Timing: reminder before they are gone
Category interest Bought or viewed a given category Content: which products to feature
Suppression No engagement in a long window, bounced, marked spam Whether you send at all

1. The engaged sending audience

This is the default audience for regular campaigns. Define it from signals that are hard to fake: clicks, site visits, added to cart, orders. Include brand-new subscribers so they get a fair chance to engage.

The window depends on how often you send and how often people buy. A store that emails three times a week can use a tighter window than one that emails twice a month. Widen it for big moments such as a product launch, and narrow it when inbox placement slips.

2. Never purchased vs. customers

The single most useful split. Non-buyers need proof: reviews, how the product works, what makes you different, your returns policy. Customers already know you. Sending them the same “Why us?” email wastes their attention.

3. One-time buyers

For most stores, the gap between one order and two is where lifetime value is won or lost. One-time buyers deserve their own content: how to use what they bought, what pairs with it, and a clear next step. Track your first-to-second order rate on this group. Our article on the first-to-second purchase rate explains how to find the products that lead to a second order.

4. Repeat customers

These people already buy without a push. Give them early access, new arrivals and service, not your deepest discounts. If you run a loyalty program, this is where tier messages belong.

5. Lapsing customers

Define “lapsing” from your own data. Look at the typical gap between first and second orders, or the product’s replenishment cycle for consumables. When a customer passes that gap by a clear margin, a reminder is timely. Much later, they belong in a win-back campaign.

6. Category or product interest

Use what people bought or browsed to choose which products appear in a campaign. Two or three broad interest groups are enough to start. You don’t need a segment for every collection.

7. Suppression

People who haven’t engaged in a long time, hard bounces and spam complaints. Stop sending them regular campaigns. A written sunset policy decides when someone moves here and whether they get a last re-engagement attempt first.

How do you define “engaged” when open rates are unreliable?

Open rates used to anchor most engagement segments. They are now a weak signal. Apple Mail Privacy Protection preloads email content, which can register an open even when nobody read the message. Segments built only on opens include many people who are not paying attention.

Build engagement on stronger signals:

  • Clicks in email or SMS.
  • On-site activity such as viewed product, added to cart or started checkout, which your ESP records when its on-site tracking is installed and the visitor is identified.
  • Orders.
  • Recent sign-up, so new subscribers get a fair window.

Keep opens as a secondary condition if you like, but never as the only one. Our guide to email deliverability on Shopify explains how engagement affects inbox placement.

How many segments is too many?

More segments mean more versions to write, more chances for overlap and smaller groups that are hard to read. A simple test: if two segments would receive the same email, they are one segment.

Some practical limits:

  • Size. A segment too small to produce a meaningful number of orders per send can’t tell you much. Merge it with a neighbor until it grows.
  • Exclusivity. When segments overlap, decide which one wins. A repeat customer who is also lapsing should get the lapsing message, not both.
  • Ownership. Every segment needs someone who knows why it exists. If nobody can explain it, archive it.

Advanced segments, such as predicted next-order date or discount sensitivity, are worth building once the core seven work and you have enough volume to see a difference.

Where does RFM, or Value × Intent, fit?

The seven segments above are a starting point. RFM scoring (recency, frequency, monetary value) refines them by ranking customers on value, so a lapsing high-spender is treated differently from a lapsing one-time bargain buyer.

Our Value × Intent (RFMB) model adds a Buying Stage layer to RFM: the furthest step a person reached in the last 30 days, from B1 (visit) and B2 (product view) through B3 (add to cart) and B4 (checkout) to B5 (purchase). Value tells you how much you can afford to spend on someone. Intent tells you when to act. A high-value customer at B3 needs a reminder, not a discount. Our RFM segmentation guide and the RFM segmentation playbook walk through the build.

Worked example: one campaign, three segments

All numbers here are illustrative, chosen to show the method.

A store plans a 20%-off weekend promotion. The list has 50,000 profiles.

Version A: send to everyone. Every profile gets the same email with the code.

Version B: segment first.

  1. Suppression (15,000 profiles): excluded. No revenue lost from people who weren’t engaging anyway, and less risk of spam complaints.
  2. Repeat customers in the engaged audience (5,000): get an early-access email with new products and no code. Many would buy at full price.
  3. Never purchased and one-time buyers in the engaged audience (30,000): get the 20%-off email, because they are the group the offer is meant to move.

Now compare the outcomes in margin terms:

Version A Version B
Profiles emailed 50,000 35,000
Orders from repeat customers 300, all discounted 250, all full price
Orders from the rest 400, discounted 400, discounted
Discount given on repeat orders (avg. order $80) 300 × $16 = $4,800 $0

In this example, Version B gets 50 fewer orders but keeps $4,800 in discount margin and sends to 15,000 fewer inactive profiles. Whether that trade is right depends on your margins, which is why the comparison should be made in contribution margin, not revenue.

To know what the segmentation itself added, keep a holdout group: a small random slice of each segment that gets the old treatment. Comparing those holdouts with the segmented groups shows the real difference, without guessing.

Checklist: building your first segments

  • [ ] An engaged sending audience defined on clicks, site activity, orders and recent sign-up, not opens alone.
  • [ ] A never-purchased segment with its own trust-building content.
  • [ ] A one-time buyer segment, with first-to-second order rate tracked monthly.
  • [ ] A repeat customer segment that receives early access instead of the deepest discounts.
  • [ ] A lapsing segment based on your own typical reorder gap.
  • [ ] Two or three broad category interest segments.
  • [ ] A suppression segment and a written sunset policy.
  • [ ] Clear rules for which segment wins when someone qualifies for two.
  • [ ] A small holdout when testing any new segment.
  • [ ] Results reported as revenue per recipient and margin per send.

FAQ

What is the most important email segment for an online store?

The engaged sending audience. It decides who receives your regular campaigns, which protects deliverability for everyone else. After that, the split between customers and people who have never bought has the biggest effect on what you say.

How often should I review my segments?

Check segment sizes monthly and the definitions quarterly. Revisit the lapsing and engagement windows whenever your send frequency, product range or buying cycle changes.

Should SMS use the same segments as email?

Use the same logic, but tighter. Text messages cost money per send and feel more personal, so SMS segments are usually smaller and focused on high intent, such as repeat customers or people who started checkout. Consent for SMS is separate from email consent.

Turn this into a plan for your store

The Free Profit Audit scans your store, sizes your biggest profit leaks and builds a 90-day plan.

Free Profit Audit