Klaviyo segmentation: six core ecommerce audiences

Short answer. Start with six dynamic segments: engaged prospects, new customers, active repeat customers, VIPs, at-risk customers, and a campaign-safe engaged audience. Build them from customer-generated events, purchase timing, value, consent, and profile data. Use lists for membership and acquisition sources, and segments for changing lifecycle states.

Klaviyo segments update as profiles meet or stop meeting conditions. That makes them suitable for lifecycle state, unlike a static export named VIP final v4 that becomes stale the day after it is created.

Klaviyo's segment conditions reference supports events, profile properties, location, predictive data, and other conditions. A segment can contain up to 100 conditions, but a good definition should be explainable without using all of them.

Lists versus segments

A profile remains in a list until removed. A profile enters or exits a segment when the definition changes or their data changes. Do not create a separate list for every behavior.

Segment 1: engaged prospects

Goal: reach people who have shown interest but have never purchased.

Example definition:

Placed Order zero times over all time

AND

Clicked Email at least once in the last 90 days OR Viewed Product at least twice in the last 30 days OR Checkout Started at least once in the last 14 days

AND

Can receive email marketing

Adjust windows to traffic and purchase cycle. Keep the OR conditions inside one grouped engagement block so one genuine signal can qualify the profile.

Use this audience for category education, buying guides, product launches, and objections. Exclude it from customer-only loyalty claims.

Segment 2: new customers

Goal: recognize the first purchase and improve product success.

Example definition:

Placed Order equals 1 over all time

AND

Placed Order at least once in the last 45 days

The window should cover onboarding and expected early use. Segment further by first product only when the post-purchase advice differs.

Use this audience for setup, care, education, support, review timing, and the next logical product. Avoid pushing a second purchase before the customer has received or used the first order.

Segment 3: active repeat customers

Goal: communicate with customers who have demonstrated repeat behavior and remain inside the category's cycle.

Example definition:

Placed Order at least 2 times over all time

AND

Placed Order at least once in the last 180 days

AND

Can receive email marketing

The 180-day value is only an example. Use order-gap data. A weekly consumable and a once-a-year product need different windows.

Use this segment for launches, replenishment, cross-sell, loyalty access, and product feedback.

Segment 4: VIP customers

Goal: recognize customer value with a rule that can be explained and maintained.

Possible entry rules include:

• Spend above a category-specific threshold.
• Order count above a frequency threshold.
• Top value percentile inside a recent period.
• Historic or predicted customer lifetime value where eligible data exists.
• A combined recency, frequency, and monetary score.

Do not use one lifetime-spend threshold forever. Inflation, assortment, and customer growth can turn the segment into most of the database or almost nobody. Review distribution quarterly. The Klaviyo VIP segmentation guide covers tier design and benefits.

Segment 5: at-risk customers

Goal: identify customers who have moved beyond their normal next-order window but are not yet long-term inactive.

Example definition:

Placed Order at least once over all time

AND

Placed Order zero times in the last X days

AND

Last order occurred before X days ago

Choose X from the product or customer cohort's order-gap distribution. Add category or first-product filters when cycles differ materially.

At-risk is not the same as unengaged. A customer may browse and click without ordering. Use a winback flow that matches the reason and value of the relationship, then move persistently inactive profiles toward a sunset process.

Segment 6: campaign-safe engaged audience

Goal: create a defensible default audience for routine promotional campaigns.

Example definition:

Can receive email marketing

AND

Clicked Email at least once in the last 90 days OR Placed Order at least once in the last 180 days OR Checkout Started at least once in the last 30 days OR Viewed Product at least twice in the last 30 days

AND

Is not in high-risk inactive segment

Do not use opens as the only signal. Apple Mail Privacy Protection can generate opens without human engagement. For large sends, expand in controlled tiers rather than jumping from a 90-day engaged group to every marketable profile.

How AND and OR change the audience

AND narrows the audience because every joined condition must be true. OR broadens it because either condition can qualify.

Compare:

Clicked in 90 days AND Placed Order in 180 days

This selects customers who did both.

Clicked in 90 days OR Placed Order in 180 days

This selects profiles who did either.

Name grouped logic in a plain-language definition before building it. Then spot-check profiles that should qualify and profiles that should not.

Event filters and product-level segments

Klaviyo lets you filter events by available top-level properties. Use the actual event payload rather than guessing a property name. Product ID, SKU, collection, category, value, and currency can differ by integration.

Before launching a product-specific segment:

1. Open recent events in Analytics > Metrics.
2. Inspect property names and values.
3. Create the segment with an exact test value.
4. Check known profiles.
5. Estimate how missing or historical data affects membership.

Nested event properties may not be available for standard segmentation. Flatten important values through the integration only when the data model and maintenance owner are clear.

Segment QA before a campaign

• Consent and channel eligibility are correct.
• AND and OR groups match the written definition.
• Time windows use the intended timezone and boundary.
• Purchase and revenue currency are consistent.
• Apple privacy opens do not qualify profiles by themselves.
• Suppressed and internal profiles are excluded.
• The estimated audience change is explainable.
• At least five included profiles were spot-checked.
• At least five expected exclusions were spot-checked.
• Campaign exclusions prevent conflicting messages.

Klaviyo's 2026 documentation notes that campaign estimated audience can differ from raw segment count because consent, suppression, and other send-time rules apply. Treat the campaign recipient estimate as the final operational view.

Naming and ownership

Use a format such as:

SEG | Lifecycle | At risk | Customer | 120d

The segment description should include:

• Business definition.
• Exact conditions.
• Owner.
• Created and last-reviewed dates.
• Campaigns and flows that depend on it.
• Known data limitations.

Archive duplicates and test segments. Klaviyo currently enforces account limits for active lists and segments, so governance also prevents operational clutter.

How to measure segment quality

Do not judge a segment only by its size. Track:

• Delivered, click, and order rates.
• Revenue per recipient and contribution.
• Unsubscribe and complaint rates.
• Repeat purchase and time to next order.
• Movement into the next lifecycle state.
• Incremental lift against a holdout when possible.

A smaller segment with clear intent can create more useful revenue and fewer negative signals than a large generic audience.

Common segmentation mistakes

• Creating lists for every behavioral state.
• Combining AND and OR without grouping intent.
• Using opens alone for engagement.
• Copying another account's time windows.
• Defining VIP by an arbitrary fixed spend only.
• Ignoring product cycle and currency.
• Building segments from event properties that are not populated.
• Keeping dozens of near-duplicates with no owner.

FAQ

Do Klaviyo segments update automatically?

Yes. Dynamic segments recalculate as profiles meet or stop meeting their conditions. Updates may not always appear instantly for every time-based condition, so verify critical send audiences before launch.

How many conditions can a Klaviyo segment have?

Klaviyo's current reference allows up to 100 conditions. That is a technical maximum, not a design recommendation. Prefer the smallest definition that expresses the business rule.

Should engagement use opens or clicks?

Use clicks and customer events as stronger signals. Opens can be included as a secondary signal when Apple privacy behavior is accounted for, but should not keep a profile active on their own.

What is the difference between at-risk and inactive?

At-risk customers recently crossed their expected purchase or engagement window. Inactive profiles have remained without useful signals for a longer period and may need a sunset path rather than another promotional sequence.

How often should segments be audited?

Review core send and suppression segments monthly, value and lifecycle thresholds quarterly, and every dependent segment after an integration or event-schema change.

Make every audience explainable

Good segmentation turns customer data into a decision the team can defend. Deliver helps ecommerce teams design the definitions, flows, campaigns, and measurement that depend on them. Book a Klaviyo and CRM diagnostic.