Email Strategy
Email Segmentation Guide: Segments That Match How Customers Buy
Email Segmentation Guide: Segments That Match How Customers Buy
Tarik Abida · September 12, 2026 · 11 min read
Most brands do not have a list problem. They have a sameness problem. One welcome. One weekly campaign. One discount. Eight thousand people who bought a $40 serum and a $280 coat get the same subject line.
That is not an email program. That is a newsletter hobby with extra steps.
This email segmentation guide is the version we use on email marketing retainers: a small set of segments that match how the customer actually buys, then flows and campaigns that refuse to treat everyone as average.
If you want the account diagnosed against this map, schedule a free audit.
Why generic email lists stall revenue
A blended open rate hides the leak. VIP buyers still open. Window shoppers still ignore you. The dashboard looks “fine” while repeat purchase is stuck.
On most Klaviyo audits we see the same four failures:
- Everyone is “Active.” The platform’s engagement cluster is not a strategy. It is a deliverability band-aid.
- Campaigns ignore last product. Someone who bought skincare gets a jacket drop because the merch calendar said so.
- Flows ignore value. Welcome is a three-email coupon. Post-purchase is a review ask. VIP does not exist.
- Unengaged people still get the Tuesday send. That trains Gmail to bury the brand, which is a deliverability problem dressed up as “our opens dipped.”
Segmentation is how you stop averaging. It is also how you protect inbox placement: high-intent people get more, cold people get less, and the domain stays out of spam.
The five segments to build first
Do not start with 40 slices. Start with five that a founder can explain on a call. Everything else is a child of these.
| Segment | Who is in it | What they get | Success metric |
|---|---|---|---|
| New subscribers | Opted in, 0 orders, last 14 days | Welcome + education, not a blast calendar | First-order rate from welcome |
| Buyers | 1+ order, not VIP | Post-purchase, replenishment, cross-sell | Repeat purchase in 60 days |
| VIP | Top 20% of revenue or 3+ orders | Early access, service, no panic discounts | VIP share of email revenue |
| At-risk | Bought once, quiet 45–90 days | Winback with a reason, not 20% off by default | Re-order rate |
| Unengaged | No open/click 90–120 days | Sunset series, then suppress | List health, not campaigns sent |
How to define VIP without a vanity club
VIP should be a revenue rule, not a badge in the footer.
Use one of these, not all of them:
- Top 20% of customers by 12-month revenue.
- Three or more orders.
- AOV above a threshold that actually matters for the catalog (do not set $50 VIP on a $180 average order).
If the rule cannot be written in one sentence, it will not get maintained.
Unengaged is a hygiene segment
Unengaged is not a campaign idea. It is how you stop paying to mail people who will never buy. A sunset series (two or three emails), then suppression. Brands that skip this slowly train the inbox to ignore the domain. That is the unsexy half of email deliverability.
How to set this up in Klaviyo
Klaviyo will let you over-build. Resist it. Name segments like an operator, not like a dashboard.
{% comment %} Example: VIP flag in a campaign {% endcomment %}
{% if person.VIP == true %}
Early access is live. 24 hours before the list.
{% else %}
The drop is Tuesday. You are on the list.
{% endif %}
Practical build order:
- Profile properties you will actually use. VIP (boolean), predicted next order date if you have it, last collection purchased. Do not create 30 properties “for later.”
- Five parent segments. New, buyers, VIP, at-risk, unengaged. These are the campaign filters.
- Flow filters, not extra segments. Welcome excludes buyers. Post-purchase excludes people already in a replenishment flow. Winback excludes VIP (they get a different tone).
- Campaigns use AND logic. “Buyers AND not unengaged AND (skincare last order OR hair last order).” If the filter needs a paragraph, it is too clever to QA.
For the full account build, that is Klaviyo agency work. For the map only, it is a Klaviyo experts session.
Campaign calendar after segments exist
Once the five segments exist, the weekly send stops being “everyone minus unsubscribes.” A normal week looks like this:
- Tuesday: buyers + VIP, product or story. No discount required.
- Thursday (optional): new subscribers who have not converted — education, not the same merch blast.
- VIP only: early access 24 hours before a real drop. If there is no real drop, do not invent one.
- Never: unengaged on the main campaign. They are in sunset or they are suppressed.
What to send each segment (and what to stop sending)
New subscribers
They opted in. They did not ask for the whole catalog.
Send a welcome that explains what you sell, who it is for, and one clear first purchase. Coupons are optional. Brand is not. If the only reason to stay subscribed is 15% off, you will re-earn that discount on every send.
Buyers
Post-purchase should do three jobs: reduce returns (how to use), create the next order (replenish or pair), and collect proof (review / UGC) after they have used the product. Asking for a review on day 1 is lazy. Asking on day 14 for a serum is a system.
See lifecycle marketing for the full journey. Segmentation is what makes that journey different per person.
VIP
VIP email should feel like the brand noticed. Early access. Stock on a size they bought. A human-sounding note when something is late. What VIP should not feel like: the same 20% off the list just got, with “exclusive” in the subject.
At-risk
Winback is not a coupon sequence with three increasing discounts. Lead with a reason they might care: new formula, restock of what they bought, a use they missed. Discount last, if at all. Luxury and high AOV lists punish panic discounting — we wrote that up in the luxury fashion case study.
Unengaged
Two or three emails. “Still want this?” Then suppress. Keeping them “for list size” is how you buy a deliverability problem.
How to measure segmentation (without lying to yourself)
Opens are a weak scoreboard, especially after Apple’s privacy protection. Use this instead:
| Question | Metric | Where it lives |
|---|---|---|
| Did welcome create buyers? | Placed order rate, welcome revenue | Welcome flow |
| Do buyers come back? | 60-day repeat rate by segment | Cohort / LTV |
| Is VIP worth the extra work? | % of email revenue from VIP | Campaign + flow |
| Is winback real? | Re-order rate vs holdout | Winback flow |
| Is the list getting healthier? | Unengaged % of total profiles | List |
Holdouts matter. If you cannot turn a segment off for 10% of profiles, you cannot prove the segment did anything.
A/B tests on subject lines are fine after the segments exist. They are a waste before. Read A/B testing email campaigns once the map is live.
A 14-day build you can actually finish
Days 1–2. Export a simple picture: orders per customer, last order date, last collection. You are looking for the 20% who pay the bills and the pile who never bought.
Days 3–5. Build the five segments in Klaviyo. Name them so a new hire understands them. Screenshot the definitions. If you cannot screenshot a definition, it is not a system.
Days 6–9. Attach flows: welcome excludes buyers; post-purchase and replenishment for buyers; winback for at-risk; sunset for unengaged. Campaigns: exclude unengaged immediately, even if flows are still ugly.
Days 10–14. One VIP-only send. One buyer send that references last collection (even if the template is simple). Measure placed orders, not star ratings on the copy.
Most of the common email mistakes we see are skipping this order and jumping to “better templates.”
Segmentation is not a feature you toggle. It is the difference between a list and a revenue system. If you want that system built in the account — not a deck — talk to SuperMailo.
Next step
Ready to improve your email strategy?
Schedule a consultation. Leave with a punch list — even if SuperMailo is not the partner.