SMS vs Email for Ecommerce: When to Use Each (2026)

If email is generating less than 25% of your store's revenue, the problem usually isn't your copy, your templates, or your send frequency. It's that you're sending the same message to a 50,000-person list where some people bought yesterday, some haven't opened anything since March, and some have spent $2,000 with you this year. Same email, wildly different relationships.

Klaviyo's own benchmark data makes the case plainly: properly segmented campaigns earn roughly 3x more per recipient than unsegmented sends. Segmentation isn't an optimization layer you add later; it's the difference between email as a top-three revenue channel and email as background noise.

Here are the Klaviyo segmentation strategies we build into every account we manage, in the order they pay off.

Start with engagement tiers: they protect everything else

Before any clever behavioral segments, build the three tiers that determine who gets your campaigns at all:

• 90-day engaged: opened, clicked, was active on site, or purchased in the last 90 days. This is your default campaign audience.
• 90–180 day lapsing: gets your best sends only: major launches, your strongest offers. Not the Tuesday newsletter.
• 180+ day dormant: campaigns stop entirely. This tier belongs to your sunset flow, which re-engages or suppresses them.

This isn't just targeting; it's deliverability protection. Sending to disengaged profiles tells Gmail your mail gets ignored, which drags down inbox placement for the subscribers who do want your email. Engagement tiers are where segmentation and ecommerce email deliverability become the same discipline.

Layer in purchase behavior: the RFM framework

Engagement tells you who's paying attention. Purchase behavior tells you who's paying. The simplest framework that captures it is RFM (recency, frequency, monetary value), and you don't need all nine boxes of the classic grid to get the revenue. Four segments do most of the work:

• VIPs (3+ orders or top ~10% by lifetime spend): early access, insider treatment, no discounts needed. These customers respond to status, and discounting them is burning margin on purchases that were happening anyway.
• Recent first-time buyers (1 order, last 60 days): the highest-leverage segment in ecommerce. The jump from first to second purchase is where most brands lose the customer. This segment gets education, cross-sells based on what they bought, and a reason to return.
• Needs attention (2+ orders, nothing in 90+ days): proven buyers going quiet. A "we noticed" send with a relevant offer here reliably outperforms any campaign to cold subscribers.
• Engaged non-buyers (active on email, zero orders): interested but unconverted. They need social proof and objection handling, not another product grid.

Each of these segments implies different messaging, different frequency, and different offers, which is exactly the point. When every segment gets the same campaign, you're leaving the differences on the table.

Use Klaviyo's predictive segments once you have the data

With enough order history, Klaviyo starts predicting customer lifetime value, churn risk, and expected next order date for every profile. That unlocks segments you can't build by hand:

• High predicted CLV, low current spend: treat tomorrow's VIPs like VIPs today, before a competitor does.
• High churn risk with purchase history: intervene while they're still reachable, instead of discovering the churn in a quarterly report.
• Approaching expected order date: time replenishment and winback sends to when the customer is actually due, not to your campaign calendar.

We covered how these models work and when to trust them in our guide to Klaviyo predictive analytics. The short version is they need order volume to be accurate, so lean on RFM until the predictions stabilize.

Exclusion segments: the half everyone skips

Who you don't send to matters as much as who you do. Three exclusions belong on nearly every campaign:

1. Recent purchasers (last 14–30 days, tuned to your reorder cycle). Nothing reads worse than a discount on the product someone bought at full price last week.
2. Active subscription customers: they don't need the pitch for something they already receive monthly; they need different messaging entirely.
3. Profiles currently in high-intent flows: someone mid-abandoned-cart-sequence shouldn't get a conflicting campaign offer that undercuts the flow.

The mistakes that quietly cap your revenue

• Building segments on opens alone. Apple's privacy features inflate open data. Anchor engagement definitions to clicks, site activity, and purchases, with opens as a supporting signal.
• Too many segments, too fast. Twenty segments nobody maintains lose to six segments with distinct messaging. Add a segment only when it changes what you'd say.
• Segmenting campaigns but not flows. A welcome series that branches for buyers vs. non-buyers, or a winback that treats VIPs differently, compounds every single day; flows are where segmentation earns the most.
• Set-and-forget definitions. Review segment sizes quarterly. If your "engaged" segment is shrinking while your list grows, that's a program problem the dashboard won't flag on its own.

FAQ

How many Klaviyo segments should an ecommerce brand actually have?

For most brands in the $1M–$50M range: three engagement tiers, four to five purchase-behavior segments, and two or three standing exclusions, roughly ten total. More than that only makes sense once each existing segment gets genuinely different messaging.

What's the difference between a Klaviyo list and a segment?

Lists are static: people join by opting in. Segments are dynamic: profiles flow in and out automatically as they meet or stop meeting your conditions. Almost everything in this post should be a segment, so the targeting updates itself.

Should I segment by demographics like age or location?

Only when it changes the offer, such as location for regional shipping or events. For most DTC brands, behavior (what someone bought, when, and how they engage) predicts the next purchase far better than who they are on paper.

Does heavier segmentation mean sending less email overall?

Usually it means sending less to disengaged profiles and more to engaged ones. Total volume often stays flat or grows; it's the distribution that changes, and that shift is what lifts both revenue per send and deliverability.

Segment for revenue, not for tidiness

The test for every segment is simple: does it change what you send, when you send it, or what you offer? If yes, it earns its place. If no, it's organizational clutter.

Want to know which segments your program is missing? Take our free email growth assessment to benchmark where your gaps are, or book a 30-minute call and we'll audit your Klaviyo segments against what your revenue data says they should be.