Klaviyo email operations
Build a Klaviyo RFM Creative Matrix Before Duplicating the Email
A Klaviyo RFM dynamic email content workflow should begin outside the template editor. Define every RFM state—including its offer, products, proof, CTA, display rule, fallback, and preview profile—in one master brief. Then implement those decisions as controlled conditional blocks. This keeps reviewers focused on the differences that matter without creating an uncontrolled collection of nearly identical emails.
Direct answer
The practical workflow
Use one creative matrix as the source of truth and one governed Klaviyo email wherever the send timing and mechanics remain the same. Give each RFM group a documented objective, offer, product treatment, proof point, CTA, show-or-hide rule, fallback, and named QA profile. Build those treatments with conditional sections or blocks, preview every named state against representative profile data, and run a live internal campaign when you need to validate tracked links or other recipient-facing behavior. Duplicate the email only when the sequence, timing, audience exclusions, or other send mechanics genuinely differ—not merely because the hero, offer, recommendation, or CTA changes.
01
1. Confirm access and choose the right level of variation
Start by deciding whether you need content variants inside one email or genuinely different campaign and flow mechanics.
Klaviyo documents six RFM customer groups: Champions, Loyal, Recent, Needs attention, At risk, and Inactive. Its RFM personalization guidance shows that the Current RFM group or Previous RFM group profile property can be used to control content. Klaviyo also states that the functionality covered by that guidance requires Advanced KDP and Marketing Analytics rather than only its standard marketing application. Confirm access in the account before your team invests in feeds, copy, or design variants. Klaviyo’s RFM personalization guide provides the relevant account caveat and implementation context.
Next, separate content differences from delivery differences. A changed hero, offer, proof block, product feed, or CTA can usually be governed as a state within one message. A different delay, sending cadence, suppression policy, trigger, or multi-email sequence is a structural difference and may justify a separate flow path or message. Klaviyo supports both RFM-driven conditional content and conditional flow paths, but they solve different operational problems.
The objective is not to force every scenario into one complicated template. It is to prevent cosmetic differences from producing uncontrolled copies. Use one email while reviewers can still understand and test its states reliably. Split the implementation when the communication strategy—not merely the content treatment—has changed.
- Verify that the required RFM properties and account features are available.
- List every proposed difference between groups.
- Mark each difference as content, merchandising, offer, timing, sequence, trigger, or suppression.
- Keep content-level differences in the matrix; create separate delivery paths only for material send-mechanic differences.
02
2. Create the master RFM creative matrix
The matrix should make every variant reviewable before anyone starts copying blocks or editing template logic.
Create one row for every RFM treatment and one additional row for missing or unexpected data. At minimum, record: RFM group, campaign objective, offer, product recommendation, proof, CTA label, CTA destination, show rule, default behavior, product fallback, preview profile, owner, and approval status. Copy profile-property names and values exactly as they appear in Klaviyo instead of paraphrasing them in the brief.
A useful schema is: RFM group | objective | offer | primary product treatment | product fallback | proof | CTA label | CTA URL | show rule | hide/default rule | preview profile | creative owner | CRM owner | QA status. If legal language, coupon terms, inventory constraints, or regional exclusions differ, add explicit columns for those items. Do not bury them in comments.
The fallback row is essential. It should define what a person sees when Current RFM group is absent, unrecognized, delayed, or otherwise unsuitable for a specialized treatment. A safe fallback is usually a broadly relevant brand message and static merchandising block that does not claim VIP status, lapse, loyalty, or a personal recommendation.
The matrix also controls approvals. Instead of asking stakeholders to approve “the email,” ask them to approve CHAMPION, LOYAL, RECENT, NEEDS_ATTENTION, AT_RISK, INACTIVE, and NO_VALUE. This exposes conflicting discounts, unsupported product promises, inconsistent landing pages, and missing states before scheduling.
- Create one row per intended RFM state and one default row.
- Enter the exact profile property and expected value for every conditional rule.
- Write the visible offer, proof, CTA, and product logic rather than using vague labels such as “personalized version.”
- Assign a named preview profile and an accountable QA owner to every row.
- Lock the matrix for final review before implementing late-stage changes in Klaviyo.
03
3. Map each RFM group to a restrained creative direction
RFM should guide the message objective, but it should not automatically dictate an aggressive discount.
Klaviyo describes its RFM groups and suggests actions for each in its RFM scoring and customer-groups documentation. Use those descriptions as a starting point, then adjust the treatment for your margins, replenishment cycle, inventory, loyalty program, and brand positioning.
Champions: Recognize the relationship through early access, premium products, loyalty benefits, or a review request. Avoid giving away margin by default when access or recognition is sufficient. Loyal: Emphasize relevant cross-sells, replenishment, rewards, or popular products that complement established buying behavior. Recent: Reduce uncertainty around the second purchase with customer proof, complementary products, onboarding education, or an appropriate second-order incentive.
Needs attention: Lead with meaningful newness, replenishment, or products related to prior purchases. If you use urgency or an incentive, document the eligibility and expiry in the matrix. At risk: Use direct win-back framing, recognizable bestsellers, or accessible products. Keep the message focused rather than stacking several offers. Inactive: Use a clear reactivation proposition and broadly appealing merchandising, then review whether continued sending is appropriate under your engagement and deliverability policies.
These are planning directions, not universal performance claims. The matrix should explain why a group receives its treatment and what would cause the team to revise it. If two groups receive the same content, document that intentionally rather than creating nominal variants with no meaningful creative distinction.
- Assign one primary communication objective to each group.
- Choose the least complex offer capable of supporting that objective.
- Select proof that is accurate and relevant to the proposed product treatment.
- Write one primary CTA with a specific destination.
- Record any margin, inventory, eligibility, or legal constraint affecting the variant.
04
4. Implement the matrix with conditional blocks
Translate each approved row into a documented display rule, keeping the template hierarchy as simple as possible.
Klaviyo’s dynamic show-or-hide documentation explains how blocks and sections can be controlled through its logic builder. For an RFM treatment, select the relevant section or block, open its display settings, choose the logic builder, and set Current RFM group to equal the exact group value recorded in the matrix.
Keep related elements together when they must change together. If a group-specific hero includes a headline, image, offer line, and CTA, place them inside one conditional section rather than applying independent rules to every child block. Nested parent-section and child-block rules can interact, making it harder to diagnose an empty or contradictory result.
Klaviyo’s conditional logic reference notes that comparisons can be case-sensitive. Copy the expected property values directly, and preview the result against profiles that actually contain those values. Do not assume that a visually similar value, capitalization change, or internal shorthand will match.
Keep RFM profile logic conceptually separate from trigger-event personalization. Klaviyo’s logic builder supports profile data types including text, numbers, and lists, while some other data types and event-data scenarios require custom-coded conditions. If a message combines RFM treatment with event-specific content, document both sources and test their interaction rather than assuming one preview proves the other.
- Add the approved block or section.
- Open Display and select the logic builder.
- Set
Current RFM group > equals > [exact group value]. - Avoid additional conditions unless the brief explains why they are necessary.
- Document the implemented rule and template location in the matrix.
- Add one unconditional or deliberately defined default treatment for unmatched profiles.
05
5. Define product logic and fallback merchandising
A product recommendation is not complete until its source, limitations, and fallback are visible in the brief.
For every RFM row, identify whether the product area uses a dynamic recommendation, a named product feed, or manually selected products. Klaviyo distinguishes dynamic product blocks from static blocks in its product-block guidance. Dynamic blocks populate from catalog and recommendation logic, while static blocks use products selected by the marketer.
Record the intended merchandising logic in plain language. “Personalized products” is insufficient. A better instruction is: “Show complementary products based on prior purchase context; if that treatment cannot render, show the approved bestseller collection.” The fallback must remain truthful for any recipient who sees it.
Klaviyo notes that product blocks select products rather than a particular product variant. This matters if the creative promises a specific shade, size, pack configuration, or SKU. The headline, image, and CTA should not imply that a precise variant will appear unless the landing experience and implementation reliably support that promise.
During QA, inspect the actual catalog results rather than approving editor placeholders. Check product names, images, prices, destination pages, availability, and the relationship between the recommendation and surrounding copy. A technically functioning block can still create a poor experience if its merchandising contradicts the offer or proof.
- Name the primary product source for every RFM state.
- Define an acceptable static or broadly relevant fallback treatment.
- Remove copy that overpromises a specific product variant.
- Preview real catalog output for every named profile.
- Verify product images, titles, prices, CTA labels, and destinations together.
06
6. Create named profiles and review every state
A conditional email is a collection of render states. Each state needs a profile, owner, and recorded result.
Create or identify representative profiles named QA_RFM_CHAMPION, QA_RFM_LOYAL, QA_RFM_RECENT, QA_RFM_NEEDS_ATTENTION, QA_RFM_AT_RISK, QA_RFM_INACTIVE, and QA_RFM_NO_VALUE. The no-value profile should omit the relevant property or contain a deliberately unmatched value so the team can verify the default.
Populate each QA profile with the properties necessary for its intended result. Klaviyo’s conditional-content guidance recommends testing across the expected conditions and missing or different data. Its preview and test documentation explains that reviewers can search for a profile and inspect the resulting message in desktop and mobile views. Event-triggered messages can also be previewed against relevant event instances.
For every profile, confirm that exactly one intended hero, offer treatment, proof block, product treatment, and primary CTA appears. Also look for artifacts created by hidden content: empty headings, stranded punctuation, doubled spacing, repeated dividers, orphaned legal text, or a mobile layout that assumes another block is visible.
Record the review result in the matrix. Screenshots can help stakeholders compare states, but they should supplement—not replace—a profile-based preview. The approved artifact should identify which data produced the screenshot so the state can be reproduced after revisions.
- Preview all six named RFM profiles.
- Preview the no-value or unmatched profile.
- Inspect both desktop and mobile rendering.
- Confirm that conditional sections do not overlap or leave empty structural elements.
- Record the reviewer, date, result, and required corrections for each state.
07
7. Run variant, product, and link QA before scheduling
Separate QA into explicit passes so a correct conditional rule does not conceal a broken offer, product block, or destination.
Variant QA: Confirm the exact RFM value, visible hero, offer, proof, product treatment, CTA, terms, and fallback. Check parent and child display rules for conflicts. Confirm that the default profile receives neutral content and that no profile receives two mutually exclusive treatments.
Product QA: Verify that the intended feed or static selection appears. Inspect product title, image, displayed price, CTA, and destination as a set. Check whether the surrounding copy implies a variant or availability that the block does not guarantee. Review the fallback with the same rigor as the personalized state.
Link QA: Click the logo, hero image, linked text, product images, navigation, and every CTA. Confirm landing pages, offer parameters, coupon instructions, preference-center links, and unsubscribe handling. Klaviyo states that standard preview and test emails do not apply click tracking or Google Analytics parameters, and that unsubscribe links, preference links, and dynamic coupons are not fully functional in that context. When those behaviors must be validated, its preview documentation recommends sending a live campaign to an internal preview list rather than relying only on a standard test email.
Do not schedule immediately after a late conditional change. Re-preview every state affected by the changed parent section, shared footer, product source, offer term, or CTA destination. A small shared-component edit can alter several RFM variants at once.
- Complete variant QA for every named profile and the default.
- Complete product QA with real catalog output.
- Complete a manual link inventory and click every linked element.
- Send a live internal campaign when tracked links or recipient-facing functions require validation.
- Recheck affected states after every material correction.
- Record final CRM, creative, merchandising, and link approval before scheduling.
08
8. Preserve the rationale after the send
Keep the matrix, implemented rules, approvals, and observations together so the next campaign starts from evidence rather than memory.
Archive the final brief with its exact RFM values, screenshots, product logic, fallback decisions, CTA destinations, QA notes, and change history. When the team revisits the campaign, it should be possible to distinguish the strategic decision from the Klaviyo implementation detail.
ATIYO can organize the roadmap, master brief, brand context, assets, iterations, and reusable learnings around this workflow. It does not connect to ad accounts, buy media, calculate ROAS, or automatically know performance. Media performance remains in the ad platform, and ATIYO preserves creative context and learnings when a user records them.
A practical setup is one roadmap item for the send, one custom brief containing the RFM matrix, and linked iterations for screenshots or revisions. Use the matrix fields as approval criteria so feedback such as “change the CTA” identifies the affected RFM state instead of creating an ambiguous template-wide request.
If you need a governed place to keep the matrix and its iterations, start ATIYO’s card-confirmed seven-day trial on the selected plan before billing begins.
- Archive the approved matrix with the final implementation.
- Record which states changed during production and why.
- Add observed results manually only when their source and measurement window are clear.
- Turn durable findings into reusable guidance for the next campaign rather than copying the prior email blindly.
Frequently asked questions
Questions about this workflow
Should every Klaviyo RFM group receive a different email?
No. Use conditional content when the send timing, trigger, and overall structure are shared. Create separate messages or flow paths when groups need materially different timing, sequence, suppression, or operational handling. Some groups may intentionally share one treatment.
What should appear when a profile has no RFM value?
Use a documented default that does not make unsupported claims about loyalty, inactivity, purchase history, or personal recommendations. Preview it with a named no-value profile before scheduling.
Can I approve the email from one Klaviyo test send?
Not safely when the message contains several conditional states. Preview every expected RFM value and the fallback. Use a live campaign sent to an internal list when you need to test click tracking, analytics parameters, dynamic coupons, or recipient-facing subscription links.
What is the minimum useful RFM matrix?
Include the exact RFM value, objective, offer, product treatment, fallback, proof, CTA label, CTA URL, display rule, named preview profile, owner, and approval status. Add inventory, margin, eligibility, and legal fields when they affect the treatment.
Does ATIYO read Klaviyo or media performance automatically?
No. ATIYO organizes creative strategy, briefs, assets, iterations, and reusable learnings. It does not connect to ad accounts, calculate ROAS, or know performance unless a user records it. Media performance remains in the ad platform.
Primary and official sources
Sources used in this guide
External product facts were checked against the organizations’ own documentation. Features can change; confirm current details before making a purchase or campaign decision.
- How to personalize product recommendations by RFM group | Klaviyo Help Center RFM properties, conditional-content use cases, flow paths, and the Advanced KDP and Marketing Analytics access caveat.
- Understanding scoring and customer groups in the RFM report | Klaviyo Help Center Definitions and suggested actions for Klaviyo’s six RFM customer groups.
- How to show or hide template blocks and sections based on dynamic variables | Klaviyo Help Center Logic-builder workflow, supported profile data, testing considerations, and interactions between section and block rules.
- Conditional logic reference for templates | Klaviyo Help Center Conditional syntax and case-sensitivity considerations.
- How to add a product block to an email | Klaviyo Help Center Dynamic and static product blocks, preview behavior, and product-level rather than variant-level selection.
- How to preview and send test emails in Klaviyo | Klaviyo Help Center Profile-based previews, desktop and mobile review, event previews, and limitations of standard test emails.
Move the plan out of scattered sheets
Run the roadmap, briefs, assets, and learnings in ATIYO.
ATIYO keeps the brand context and production decisions connected. It does not buy media, connect to ad accounts, or invent performance results.