Email and SMS in 2026: What Klaviyo Proved, What It Costs, and When to Add SMS

Two companies ran the same play in the same market. One of them is now worth billions and the other sold its messaging customers to a competitor and cut a third of its staff.

Klaviyo picked one job, ecommerce email, and went deep enough on it that switching became painful. Yotpo picked breadth, bundled reviews and loyalty and messaging together, and eventually sunset the email and SMS products entirely, transitioning that customer base to Attentive^[1].

That's the most useful thing in this article and it isn't about which tool to buy. Depth in one job beats breadth across five, in a category where the buyer can switch one tool at a time. Keep it in mind while you read the pricing.

What the Klaviyo numbers actually say

Klaviyo publishes real figures, so we can inspect the business rather than the marketing.

Q1 2026: revenue of $358 million, up 28% year on year. Total customers over 196,000. Customers generating over $50,000 of annual recurring revenue: 4,175, up 38% year on year. Net revenue retention 110%, up two points. Non-GAAP operating margin above 16%, the highest in the company's history^[2].

Read those two customer numbers together

196,000 customers. 4,175 of them at $50k or more. That's about 2% of the logo count carrying a very large share of the revenue.

The 194,000 aren't charity. They're distribution, reviews, agency familiarity, integration demand and a hiring pool of people who already know the product. But the money comes from a small mid-market cohort, and the product is priced so that growth automatically moves you toward it.

Which brings us to the thing you're actually here for.

The pricing model is the product

Every tool in this category prices on a number that grows without you doing anything.

Tool	Priced on	Pricing published?
Klaviyo	Active profiles, plus SMS credits[3]	Yes, with a calculator
Mailchimp	Contacts and send volume[4]	Yes
Omnisend	Contacts, with email and SMS tiers[5]	Yes
Attentive	Quote-led, messaging-first[6]	Largely no
Shopify Email	Free allowance then per-email[7]	Yes

The word "profiles" is doing a lot of work in that first row. A profile is a person in your database, whether or not you're emailing them, whether or not they ever bought. Which means your bill grows with the size of your list, not the value you get from it.

That's not a scam. It's a real cost driver for the vendor and it aligns roughly with your business size. But it does produce a specific behaviour that almost nobody does and everybody should.

The suppression habit

Somebody has to be responsible for removing profiles who have not opened anything in twelve months. Not archiving them mentally. Removing them from the billable count.

Every store I've looked at has between 20% and 50% of its list in this state, and every one of them was paying full price for it while it quietly damaged deliverability. Two problems, one fix, and it takes an afternoon.

I ran a list for four years without ever doing this. Open rates drifted down and I blamed the creative.

Does this actually impact you? The order-count test

Under about 100 orders a month

Use whatever's free. Shopify Email includes a monthly allowance and then charges per email, which for a small list is close to nothing^[7]. Mailchimp and Omnisend both have free tiers^[4][5].

At this size, the flows matter and the tool doesn't. Set up a welcome sequence, an abandoned cart, and a post-purchase. That's most of the value in the category and it's available for free everywhere.

100 to 1,000 orders a month

This is where a proper ecommerce ESP earns its money, because segmentation on purchase behaviour starts to be worth something and generic email tools are bad at it. Klaviyo's advantage here is real and it's mostly about the data model rather than the editor.

Over 1,000 orders a month

You're now the customer the pricing was designed for, and you should be negotiating rather than self-serving. Also the point at which SMS becomes worth the compliance overhead.

When you should not add SMS

If your repeat purchase rate is under about 20%, SMS will not fix it. SMS is a frequency amplifier on an existing relationship, and amplifying a relationship that doesn't exist gets you complaints and carrier filtering. Fix the product or the post-purchase experience first.

What we found: the free tier is better than the category admits

We compared what each vendor publishes for a store with 5,000 contacts and moderate send volume, using only their own pricing pages^[3][4][5].

Finding one: the gap is segmentation, not sending

Every tool sends email competently. What differs is what you can segment on. A generic ESP segments on list membership and email engagement. An ecommerce ESP segments on what somebody bought, when, how often, for how much, and what they browsed.

If your marketing plan doesn't currently use any of those, you're paying an ecommerce premium for a generic use case. That's a real and common waste.

Finding two: the switching cost is the flows, not the list

Exporting contacts is trivial. Rebuilding fourteen automated flows with their branching logic, timing and templates is a fortnight of somebody's life. Vendors know this. It's why the entry tiers are cheap and the renewal conversations are not.

Before you build flow number fifteen, ask whether it's earning its keep, because each one raises your switching cost.

Finding three: SMS pricing is genuinely hard to compare

Email pricing is a monthly number. SMS pricing is credits, per-segment charges that vary by country, carrier fees, and short code or 10DLC registration in the US. Two vendors quoting "SMS included" can differ by several multiples on the same volume. Model your own country mix rather than trusting a comparison table, including the one above.

Deliverability, which is upstream of everything else

You can pick the best tool in the category and still land in spam. Three things decide that, and none of them are features you buy.

Authenticate your sending domain

SPF, DKIM and DMARC on the domain you actually send from. The major mailbox providers tightened bulk-sender requirements and unauthenticated mail from a bulk sender now gets filtered rather than delivered. Every ESP documents its own setup and it's a one-off job of an hour or two.

Send from a subdomain, not your root domain

Marketing mail from a subdomain keeps a reputation problem in marketing away from your order confirmations and password resets. Getting transactional mail caught in a marketing reputation dip is a genuinely bad afternoon and I've had it.

Engagement is the ranking signal

Mailbox providers weight recent engagement heavily. Which is why the suppression habit above pays twice: your bill drops and your inbox placement improves. Sending to a list where 40% haven't opened anything in a year actively damages delivery to the 60% who do.

If your open rates have drifted down over eighteen months with no change in creative, this is usually why, and no amount of subject-line testing will fix it. Related reading: the analytics stack piece on measuring any of this honestly.

The compliance layer that gets skipped

This is where people get into genuine trouble, and it's boring enough that it gets delegated to nobody.

UK and EU: PECR and the soft opt-in

Electronic marketing to individuals generally requires consent. The exception, the soft opt-in, is narrower than people think: you may email an existing customer about similar products where you collected the address in the course of a sale or negotiations for a sale, and you gave a simple opt-out at collection and in every message^[8]. The underlying rule is Regulation 22 of PECR^[9].

Three things break this in practice. Competition entrants aren't customers. Abandoned-cart addresses are arguable at best, since no sale or negotiation completed. And "similar products" means similar, not "anything we sell".

US: CAN-SPAM

Accurate headers, honest subject lines, a working opt-out honoured within ten business days, and a physical postal address in every message^[10]. Lighter-touch than PECR, and the penalties are per message, which adds up faster than people expect.

SMS is stricter everywhere

Express consent, no soft opt-in equivalent, and a separate consent from email. Bundling SMS consent into an email signup checkbox is the single most common mistake in this category. Vendors document their own requirements and both Mailchimp and Omnisend set out the consent rules for their SMS products^[11][12].

AI-written and AI-personalised messages

Every vendor in this category shipped AI subject lines, AI copy generation and AI send-time optimisation in the last two years. Content generation is largely unproblematic. What's worth a look is anything that profiles or scores individuals to decide what they see, because that carries transparency and data-protection obligations that a marketing team won't have considered. Our EU AI Act compliance guide for ecommerce covers what genuinely applies and what's noise.

If you are on Shopify

Start with what's included

Shopify Email is in-plan with a monthly free allowance, then priced per additional email^[7][13]. For a store under a few hundred orders a month it covers campaigns adequately. Where it's weaker is complex branching automation, which is exactly what you're paying an ESP for.

The integration is the reason Klaviyo wins here

Deep Shopify integration means order, product and browse data arriving without you building anything, via the Admin GraphQL API^[14]. That's the moat. It's also why the same product is noticeably less impressive on other platforms.

Watch the profile count after a sale

A Black Friday spike adds thousands of one-time buyers to your billable profile count and most of them never open anything again. Set a calendar reminder for February to suppress the non-engagers^[3].

Not on Shopify? The other platforms

WooCommerce

The honest position: ecommerce ESP integrations on Woo are real but shallower, because the data model varies by plugin stack. Mailchimp maintains an official WooCommerce integration^[15], and MailPoet runs inside WordPress itself, which suits stores that want sending without another subscription^[16].

The specific thing to check on Woo is whether your ESP reads orders through the REST API rather than querying the database, because High-Performance Order Storage moved where orders live^[17][18]. Older integrations that query directly are broken or heading that way.

Magento and Adobe Commerce

Native transactional email templates and marketing tools are solid, and the platform expects you to bring your own marketing automation^[19]. Adobe will sell you its own marketing stack, which is capable and priced for enterprise, and often needs a separate licence from Commerce itself. Worth pricing the third-party route in parallel.

BigCommerce

Connectors to the major ESPs, clean order APIs to build against^[20]. Fewer options than Shopify, decent quality among the ones that exist.

Headless

You're sending events yourself, which is more work and better outcomes, because you control exactly what a profile knows. Get the event schema written down before you build flows against it. Headless stores are undercounted in every platform statistic you'll read^[21].

The flow audit: a table worth stealing

Score each flow. Anything with no revenue attributed in 90 days gets turned off, not improved.

Flow	Runs?	Revenue, 90 days	Verdict
Welcome series			Usually the highest earner. Fix first
Abandoned cart			Check the consent basis in the UK and EU[8]
Abandoned browse			Weaker than cart, often not worth the creep
Post-purchase			Where review requests belong
Replenishment			Only for consumables. Enormous when it fits
Winback			Test against simply suppressing them
Back in stock			High intent, low volume, always keep
Price drop			Trains people to wait. Use sparingly
Birthday			Almost always underperforms its reputation
SMS: order updates			Transactional, high value, low risk
SMS: promotional			Needs separate express consent[11]

My stance, and people do argue with it: the birthday email is the most sentimentally defended and least productive flow in ecommerce. Turn it off for a quarter and see whether anything changes.

What agents change about owned messaging

Something quietly important here that I don't see discussed.

If a customer buys through an AI assistant, they may never see your storefront, which means they may never see your signup form. The consent capture point moves from a page you control to a checkout you don't. Under the emerging commerce protocols, the merchant remains the merchant of record and the contact data flow is defined by the protocol rather than by your popup^[22][23].

Practical consequence: your post-purchase email is the surviving relationship-building surface for agent-mediated orders, because it's the one thing that reaches the buyer regardless of how they got there. That makes it more valuable than it currently is in most stores, where it's a receipt with a logo on it.

Shopify reported AI-referred orders up roughly 13 times year on year in Q1 2026, from a small base^[24]. Not urgent. Worth knowing where it's heading.

What to do this week

• Count your billable profiles against your engaged ones. Anyone with no open in 12 months is costing you money and deliverability^[3].
• Pull 90-day revenue per flow. Turn off anything at zero.
• Check your abandoned-cart consent basis if you market to the UK or EU^[8].
• Check SMS consent is collected separately from email. One checkbox for both is the common failure^[11].
• Verify a postal address appears in every campaign if you send to the US^[10].
• Read your welcome series as a customer would. It's usually your best-performing flow and usually two years old.
• Price the tier above yours at your projected list size in twelve months, before you renew.

The takeaway

If your list is small because your traffic is small, this article is the wrong problem. Go and read the channel piece instead.

Klaviyo built a $358-million-a-quarter business by being excellent at one job for one kind of customer, with 196,000 logos supplying distribution and 4,175 supplying the revenue^[2]. The suite that tried to bundle messaging with everything else exited messaging^[1].

For your store the lesson is inverted. You don't need the best tool. You need the flows working, the list clean, and the consent basis defensible, and none of those three depend on which logo is in your admin.

Most stores would make more money from suppressing dead profiles and rewriting a two-year-old welcome email than from any migration.

When did you last read your own welcome sequence, start to finish, as a customer?

Sources

1. Yotpo, press room. Vendor source for the Email and SMS sunset and the Attentive transition. Accessed 22 July 2026.
2. Klaviyo, "Klaviyo delivers strong Q1 2026 results", 5 May 2026. Revenue $358M, 196,000+ customers, 4,175 customers over $50k ARR (+38% YoY), NRR 110%.
3. Klaviyo, "Pricing". Accessed 22 July 2026. Vendor page.
4. Mailchimp, "Marketing pricing". Accessed 22 July 2026. Vendor page.
5. Omnisend, "Pricing". Accessed 22 July 2026. Vendor page.
6. Attentive, "Pricing". Accessed 22 July 2026. Vendor page; largely quote-led.
7. Shopify, "Shopify Email". Accessed 22 July 2026.
8. Information Commissioner's Office, "Electronic mail marketing", Guide to PECR. UK regulator. Accessed 22 July 2026.
9. UK legislation, "PECR 2003, Regulation 22". Primary statutory text.
10. Federal Trade Commission, "CAN-SPAM Act: a compliance guide for business". Accessed 22 July 2026.
11. Mailchimp, "About SMS marketing", including consent requirements. Accessed 22 July 2026.
12. Omnisend, "SMS marketing". Accessed 22 July 2026. Vendor page.
13. Shopify, Shopify Email app listing. Accessed 22 July 2026.
14. Shopify, "Admin GraphQL API". Accessed 22 July 2026.
15. WooCommerce, "Mailchimp for WooCommerce". Accessed 22 July 2026.
16. WordPress, "MailPoet" plugin directory listing. Accessed 22 July 2026.
17. WooCommerce, "High-Performance Order Storage". Accessed 22 July 2026.
18. WooCommerce, "REST API". Accessed 22 July 2026.
19. Adobe, "Marketing menu". Adobe Commerce documentation. Accessed 22 July 2026.
20. BigCommerce, "Orders API". Accessed 22 July 2026.
21. HTTP Archive, "Web Almanac 2025: Ecommerce". Accessed 22 July 2026.
22. Google, "Under the Hood: Universal Commerce Protocol (UCP)". Google Developers Blog.
23. Agentic Commerce Protocol, specification repository. Maintained by OpenAI and Stripe.
24. Shopify, financial reports, Q1 2026. AI-referred orders up roughly 13x YoY.