Omnichannel Email, SMS, and WhatsApp Marketing: What Esmi’s 16X ROI Claim Really Teaches Brands

Omnichannel email SMS WhatsApp marketing can create a powerful retention engine—but only if every message is relevant, consented to, deliverable, and measured against the right baseline. A recent Esmi Skin Minerals testimonial makes that point clearly: the brand says its lifecycle program has become a major returning-revenue driver, yet the more useful lesson is the operating model behind the headline.

In the original Esmi Skin Minerals interview supplied for this article, representatives describe Klaviyo as central to their retention strategy across email, SMS, and WhatsApp. They say 90% of returning-customer revenue is attributed to the platform and report a 16X ROI from SMS and WhatsApp activity. Those are impressive self-reported figures, but they should not be read as universal benchmarks—or as proof that simply turning on three channels produces a 16X return.

The real takeaway is more durable: when a brand unifies customer data, permission records, behavioral signals, message orchestration, and revenue measurement, it can stop treating email, SMS, and WhatsApp as separate campaign calendars. Instead, it can decide which message a particular person should receive, on which channel, and at what moment.

The Esmi Skin Minerals story: strong results, carefully interpreted

Esmi Skin Minerals is an Australian beauty brand that has built a retention-oriented CRM program around customer journeys rather than one-off sends. In the testimonial, Group CRM Manager Carlie Campbell and Head of Digital Tanya describe using email, SMS, and WhatsApp within one platform to tailor experiences to customer needs and channel preferences.

That narrative aligns with a published Klaviyo case study about Esmi’s earlier migration from Emarsys. The case study says Esmi adopted Klaviyo email and SMS in 2021 after dealing with incomplete customer data, limited segmentation, and manual movement of information between systems. It reported that Klaviyo drove more than 50% of online revenue in 2023, alongside a 45% customer retention rate attributed to Klaviyo. (klaviyo.com)

The apparent difference between “more than 50% of total online revenue” and “90% of returning-customer revenue” is not necessarily a contradiction. They describe different denominators:

• Total online revenue includes both new and returning customers.
• Returning-customer revenue includes only revenue from buyers who have purchased before.
• Revenue attributed to a platform depends on the attribution window and model selected.
• ROI additionally requires a definition of costs, not just attributed revenue.

This distinction matters. A marketer may hear “90% of returning revenue” and assume the platform created 90% of all repeat purchases. More precisely, it likely means that a configured attribution model credited the platform’s messages or flows with a large share of revenue generated by repeat customers. That is still meaningful, but it is a measurement outcome—not a causal proof that every purchase would disappear without the platform.

Why omnichannel email SMS WhatsApp marketing works when it works

The best omnichannel programs do not send the same copy through every available channel. They use a common customer profile to coordinate distinct channel roles.

Email is generally the high-context channel: it can carry ingredient education, routines, receipts, reviews, product comparisons, replenishment reminders, and visual storytelling. SMS is often better for concise, time-sensitive communication: a restock, early access window, cart reminder, or shipping exception. WhatsApp can add interactive and conversational experiences where the customer has explicitly chosen it, especially in markets where it is an everyday commerce and service channel.

Klaviyo now positions WhatsApp as a native channel connected to the same customer profile, flows, segments, and integrations as email and SMS. Its product materials say channel interactions synchronize to a unified profile, while its current help documentation includes segments such as highly engaged WhatsApp subscribers, high-intent shoppers, VIPs, and people who prefer WhatsApp. (klaviyo.com)

The platform is not the strategy, though. A unified profile only creates value when a brand uses it to reduce duplication and improve relevance.

One customer, one journey—not three disconnected calendars

Consider a returning skincare shopper who viewed a replenishment product but did not buy:

1. An email can explain why the product fits their past routine and include useful educational content.
2. If the customer does not engage and has separately consented to SMS, a short reminder can follow after an appropriate delay.
3. If they prefer WhatsApp and have granted WhatsApp permission, a template-based message can offer product support, stock information, or a quick route back to the cart.
4. The moment the customer purchases, every promotional branch should stop and the post-purchase branch should begin.

The final step is the one many teams miss. Without real-time purchase suppression, the customer may receive a “you left something behind” message after buying. That is not merely awkward; repeated orchestration failures train customers to ignore messages, complain, or opt out.

Channel preference is a signal, not an excuse to over-message

A customer who clicks SMS links frequently may prefer SMS, but that does not mean they want every campaign as a text. Likewise, an email subscriber may open every message and still find daily sending excessive.

A practical preference model combines at least four inputs:

• Explicit channel consent and unsubscribe status.
• Engagement recency, such as recent clicks or replies.
• Purchase behavior, including category affinity, reorder cadence, and lifetime value.
• Contact pressure, meaning how many messages the customer has received across all channels in a recent period.

The goal is not maximum reach per campaign. It is maximum useful contact over the life of the relationship.

Segmentation is the real engine behind the headline

Esmi’s representatives specifically identify deeper segmentation as a future opportunity. That is a revealing detail because segmentation is where an omnichannel program shifts from broad promotional broadcasting to retention infrastructure.

Basic segments—new subscriber, customer, non-customer, VIP—are a reasonable start. But mature segments are built around intent, eligibility, and expected next action. In beauty and skincare, that can mean skin concern, product routine, time since last order, sampling behavior, shade history, subscription status, and engagement across channels.

Klaviyo’s segmentation reporting is based on a large dataset, with its segmentation benchmark resource describing analysis of more than 250 billion emails. The broad lesson is not that any one vendor’s averages should become a target; it is that engagement and conversion should be compared across meaningful cohorts rather than judged at a whole-list level. (klaviyo.com)

Segments that create better customer experiences

For a skincare brand, useful segments might include:

• First-time buyers who purchased a starter product: Educate, gather product feedback, and introduce a complementary routine rather than immediately pushing a sitewide discount.
• Customers nearing likely replenishment: Use actual reorder patterns by SKU, not a generic 30-day timer.
• High-value repeat customers: Offer early access, concierge-style support, or loyalty recognition without making them requalify through a discount.
• Recent purchasers: Suppress them from acquisition-style messages and focus on delivery, use, review, and replenishment content.
• Dormant but historically valuable buyers: Run a tailored win-back sequence that reflects their former category interest and avoids pretending they are a new subscriber.
• Low-engagement subscribers: Reduce cadence, ask for preferences, or sunset them before they damage inbox placement.

A segment should always answer a business question. “Everyone who has not bought in 90 days” may be easy to create, but it combines a one-time trial buyer, a formerly loyal customer, a subscriber whose usual product lasts six months, and a person who never received the original messages in the inbox. Those need different treatment.

Better segmentation requires better event data

The data model matters as much as the campaign builder. A commerce platform should reliably pass events such as product viewed, added to cart, checkout started, order placed, order refunded, subscription canceled, and item out of stock.

Brands also need first-party profile properties that explain customer context. For Esmi-style skincare journeys, these may include stated skin goals, preferred products, consent source, country, language, or whether a customer has completed a quiz. Teams should only collect data they can explain, protect, and use responsibly.

For developers, the implementation question is straightforward: can your storefront, subscription platform, customer-support tool, loyalty program, and sending platform agree on customer identity and event timing? If not, marketing automation becomes an unreliable interpretation layer sitting on top of conflicting systems. Review your event schema, identity resolution rules, and webhook retries before adding more flows.

A 16X ROI claim is useful only when the math is visible

Esmi’s stated 16X ROI on SMS and WhatsApp is a valuable directional result, but a responsible team should ask how the calculation works. ROI is often used casually when people actually mean revenue-to-spend ratio or return on ad spend.

A simple revenue multiple is:

Attributed revenue ÷ channel cost

If a campaign produces $16,000 in attributed revenue and costs $1,000 in messaging fees, that is a 16X revenue multiple. It is not necessarily a 1,500% profit ROI because it does not include cost of goods, discounts, labor, agency fees, creative production, software, refunds, or the sales that would have happened without a message.

A fuller ROI calculation is:

(Incremental gross profit - total program cost) ÷ total program cost

The word incremental is crucial. A cart reminder frequently receives credit for an order, but some customers would have returned and purchased anyway. That does not make the flow worthless; it means attribution should be treated as a management tool, not an unquestionable causal answer.

Questions to ask before presenting channel ROI

When bringing results to finance or leadership, document these points:

1. Attribution model: Was revenue credited on click, open, delivered message, last touch, or multi-touch basis?
2. Attribution window: Is an order counted for one day, three days, seven days, or longer after a message?
3. Cost base: Does cost include platform fees, SMS/WhatsApp message fees, personnel, discounts, creative, and agency support?
4. Margin: Is the measurement based on revenue or gross profit?
5. Incrementality: Was there a holdout group, geo test, audience split, or other method to estimate what would have happened without the message?
6. Customer impact: Did opt-outs, spam complaints, customer-support contacts, or refunds rise as volume increased?

A strong reporting model shows both attributed revenue and a confidence-qualified estimate of incremental profit. The first helps optimize flows quickly; the second protects the business from spending decisions built on inflated credit.

Deliverability is the constraint that turns strategy into reality

A beautifully segmented lifecycle program still fails when the messages go to spam, land in a promotions tab without engagement, or are rejected before delivery. The move from one-off campaigns to always-on orchestration increases the importance of sender reputation because automated flows continuously send from the brand’s domains.

Google’s sender guidelines require all senders to Gmail to use SPF or DKIM. Bulk senders—those sending roughly 5,000 or more messages daily to personal Gmail accounts—must use SPF, DKIM, and DMARC, keep spam rates below 0.30%, support one-click unsubscribe for applicable marketing mail, and meet other technical requirements. Google also says bulk-sender classification does not expire once assigned. (support.google.com)

Yahoo similarly recommends permission-based sending and strongly urges SPF, DKIM, and DMARC for sending domains; its guidance warns against purchased lists and automatic pre-checked opt-in boxes. (senders.yahooinc.com)

The deliverability checklist for omnichannel teams

The foundational checklist is not glamorous, but it is non-negotiable:

• Authenticate every marketing sending domain with SPF and DKIM.
• Publish DMARC, begin monitoring reports, and move toward an enforcement policy only after all legitimate senders align.
• Separate marketing and transactional streams so receipts, password resets, and shipping notices do not inherit the reputation risks of promotional mail.
• Use clear From names and stable domains rather than constantly rotating identity.
• Honor unsubscribes promptly and keep suppression records synchronized across systems.
• Remove or suppress hard bounces and address repeated soft bounces according to a defined policy.
• Ramp large new sending volumes gradually rather than blasting a dormant list.
• Monitor complaint rate, authentication pass rate, bounces, deferrals, inbox placement, and engagement by mailbox provider.

Google’s current subscription guidance also says senders should implement one-click unsubscribe for subscription messages and honor requests within 48 hours. It distinguishes transactional messages, such as receipts, password resets, and one-time passwords, from subscription mail. (support.google.com)

This is why marketing and product-email infrastructure should not be treated as the same workload. Product teams need high-reliability transactional delivery and clear event handling. Marketing teams need consent-aware segmentation, campaign throttling, suppression controls, and reputation monitoring. Both need authenticated domains and clean data.

Before a record enters a marketing audience, validate it where possible with a free address verification tool. Verification does not replace permission or engagement management, but it can catch malformed and risky addresses before they create unnecessary bounces.

Consent and suppression must work across every channel

Omnichannel does not mean one consent checkbox gives a brand universal permission. Email, SMS, and WhatsApp are separate communication channels with different expectations, platform rules, and legal considerations. A customer who signs up for a newsletter has not necessarily agreed to promotional text messages or WhatsApp marketing.

For U.S. robotext programs, the FCC has emphasized that consumers can revoke consent through reasonable means and that marketers must honor revocations. Its 2024 order explains that consent revocation must be simple and that requests must be processed as soon as practicable, with a maximum of 10 business days under the rules discussed in the order. (docs.fcc.gov)

The operational implication is simple: do not store consent as a vague Boolean field called marketing_opt_in. Store the channel, capture time, capture method, disclosure version, source page, jurisdiction when relevant, and current status. Treat a STOP reply, unsubscribe action, support request, or consent revocation as an event that has to update downstream systems promptly.

Build a suppression service, not a spreadsheet

A growing company often begins with suppressions inside individual tools. That breaks down once it has separate systems for support, marketing, transactional email, SMS, and WhatsApp.

At minimum, establish a source of truth for:

• Global marketing suppression status.
• Channel-specific opt-in and opt-out status.
• Hard bounce and invalid-address status.
• Complaint and abuse signals.
• Customer-service exceptions, such as an account under review.
• Transactional eligibility, which should be narrowly defined and not used to bypass marketing opt-outs.

Your messaging systems should query or receive this information before sending. Webhooks should be idempotent, retries should not reintroduce unsubscribed users, and audit logs should make it possible to explain why a message was sent.

The best flow architecture uses channel sequencing, not channel duplication

When teams add SMS or WhatsApp, a common first move is to copy every email automation into the new channel. That can produce quick revenue, but it also creates a noisy system that turns a customer event into multiple near-identical interruptions.

A better approach uses a primary channel, a fallback channel, eligibility gates, and global frequency caps. For example, the system can lead with email for an educational welcome journey, use SMS for an expiring early-access offer only when SMS consent exists, and use WhatsApp for customer-initiated or preference-based conversations where permitted.

A practical lifecycle flow map

Here is a sensible starting architecture for a direct-to-consumer brand:

The table is a blueprint, not a prescription. A low-frequency premium skincare brand may use longer intervals than a fast-moving consumables brand. A customer who has engaged on WhatsApp may reasonably receive a different sequence than a customer who only wants email. But each branch should have a clear reason to exist.

Loyalty becomes more valuable when it is connected to lifecycle data

Esmi’s team also calls out loyalty as a future area of focus. That makes sense because loyalty is not merely a points ledger. It is a structured way to recognize a customer’s relationship with the brand and to give marketing automation better signals about value and intent.

A loyalty program can provide events such as points earned, tier reached, reward redeemed, referral completed, birthday captured, and VIP status changed. Those events can enrich post-purchase, early-access, replenishment, and win-back programs.

The risk is making loyalty another discount machine. If every loyalty event immediately triggers a coupon, the program may encourage customers to wait for incentives and can dilute margins. Better uses include education, recognition, community access, samples, routine consultations, priority support, and early product discovery.

Measure loyalty beyond redemptions

A meaningful loyalty measurement framework includes:

• Repeat purchase rate and time to second purchase.
• Customer lifetime value by loyalty tier.
• Gross margin after rewards and discounts.
• Retention compared with similar non-members.
• Redemption behavior and reward liability.
• Channel engagement and unsubscribe rates among members.
• Incremental lift, ideally measured with a holdout or matched cohort.

That framework connects directly to the Esmi lesson: retention performance improves when customer data is unified, but only if the business avoids confusing correlated engagement with incremental value.

What developers and operations teams should build behind the scenes

Marketing teams experience omnichannel orchestration as a flow builder. Developers experience it as identity, events, APIs, data contracts, webhooks, deliverability, and failure handling.

The most reliable systems make those responsibilities explicit. Your storefront should emit canonical commerce events. Your customer-data model should define how anonymous visitors become known profiles. Your consent service should keep channel permissions current. Your sending infrastructure should distinguish transactional from marketing mail and provide durable delivery-event data.

A developer-friendly implementation plan looks like this:

1. Define a canonical customer ID and document identity merge rules.
2. Standardize event names, payload fields, timestamps, and idempotency keys.
3. Send server-side conversion events in addition to browser events where practical.
4. Store consent evidence separately for email, SMS, and WhatsApp.
5. Implement purchase and unsubscribe webhooks that update flow eligibility quickly.
6. Route bounces, complaints, and delivery failures into suppression logic.
7. Keep transactional email on an API-driven, observable sending path with clear retry behavior.
8. Create dashboards for message volume, delivery outcomes, complaints, attributed revenue, and incremental-test results.

For teams building the transactional layer alongside CRM, use the email API reference and setup guides to separate application notifications from promotional lifecycle programs. The underlying principle is to avoid letting a marketing platform become an undocumented dependency for critical product communications.

The community reaction may be quiet, but the operational debate is not

There were no substantive public comments included with the source video, so there is no comment thread to treat as representative feedback. That absence is worth noting: customer-testimonial videos are naturally designed to communicate outcomes, not expose the implementation tradeoffs, attribution assumptions, deliverability problems, or compliance work behind them.

The broader industry debate is more nuanced. Marketers want unified customer profiles and fewer manual exports. Deliverability specialists warn that more available channels can lead to more contact pressure. Finance teams want credible incremental return, not simply larger attributed-revenue figures. Developers want integrations that remain dependable when APIs time out, identities merge, or a customer changes preferences.

All of those concerns can be true at once. The answer is not to avoid omnichannel marketing. It is to make orchestration deliberately constrained: permissioned audiences, channel-specific eligibility, real-time suppression, global frequency limits, and transparent measurement.

A practical scorecard for evaluating your own program

Before adopting the Esmi-style model—or expanding an existing one—assess the program across five dimensions.

1. Data readiness

Can you identify a customer consistently across store, support, loyalty, and messaging tools? Are purchase and refund events reliable? Are preference fields current enough to support real segmentation?

2. Consent maturity

Do you have clear proof of email, SMS, and WhatsApp permissions independently? Can opt-outs propagate quickly? Can support agents see the same preferences that automations use?

3. Deliverability health

Are SPF, DKIM, and DMARC aligned? Do you know your spam complaint rate, hard bounce rate, Gmail reputation signals, and mailbox-provider-specific performance? Have you separated transactional and promotional streams?

4. Journey design

Does every flow have entry criteria, exit criteria, purchase suppression, a frequency policy, and an owner? Are you sequencing channels or duplicating them?

5. Measurement credibility

Can you explain the attribution model in one sentence? Do you report revenue separately from profit? Have you run holdouts or other tests to understand incrementality?

A weak score in any one category limits the upside of the others. Great creative cannot fix a broken consent sync. Sophisticated segmentation cannot compensate for poor inbox placement. Strong attributed revenue is not enough if the same program drives complaints and erodes customer trust.

Conclusion: the durable lesson is orchestration, not a vendor claim

Esmi Skin Minerals’ reported results are compelling because they describe a brand that sees retention marketing as an interconnected system. The claim that Klaviyo accounts for 90% of returning-customer revenue and delivers a 16X return on SMS and WhatsApp should be understood as a brand-specific, self-reported outcome shaped by its data, audience, timing, attribution settings, and cost model.

What other brands can reliably learn is more practical. Bring customer data and permissions together. Segment around real needs and lifecycle stages. Give each channel a distinct job. Stop messages immediately when customer behavior changes. Protect deliverability through authentication, hygiene, and suppression. And measure performance with enough rigor to distinguish credited revenue from truly incremental profit.

That is how omnichannel email SMS WhatsApp marketing becomes a retention advantage rather than a louder campaign calendar.

FAQ

What is omnichannel email SMS WhatsApp marketing?

It is a customer-engagement approach that coordinates email, text messages, and WhatsApp around a shared customer profile. Rather than sending the same promotion everywhere, it uses consent, behavior, preferences, and lifecycle stage to select an appropriate channel and message.

Is a 16X SMS and WhatsApp ROI realistic?

It can be possible for a specific brand and measurement method, particularly for high-intent automated flows. But it should not be assumed to be a universal benchmark. Confirm whether the figure means attributed revenue divided by messaging cost or a fuller calculation based on incremental gross profit and total program cost.

Can I use one opt-in for email, SMS, and WhatsApp?

No. Treat permission as channel-specific. Maintain separate consent records for email, SMS, and WhatsApp, disclose what a person is signing up for, and make revocation or unsubscribe actions propagate promptly.

What are the most important email deliverability requirements?

Authenticate your sending domains with SPF and DKIM; publish DMARC; keep complaint rates low; support easy unsubscribes; remove invalid addresses; and separate marketing from transactional email. Gmail requires SPF or DKIM for all senders and SPF, DKIM, and DMARC for bulk senders. (support.google.com)

Should WhatsApp replace email or SMS?

Usually, no. WhatsApp is most effective as a complementary, permissioned channel with a defined role, such as interactive support, preference-led promotions, or high-intent follow-up. Email remains better for deeper content and durable records, while SMS can be effective for concise, urgent updates.