Email Marketing Strategy for Ecommerce: Flows First

Most ecommerce email programmes fail in the same order: someone turns on a handful of flows, then spends the next year sending campaigns into a list that was never segmented and never protected from decay. An email marketing strategy for ecommerce is not a content calendar. It is a build order — which flows exist, in what sequence, what segmentation sits on top of them, and what protects the list before campaign volume scales into it. Get the order right and Klaviyo’s own benchmark data, drawn from more than 183,000 brands, puts 41% of email revenue behind automated flows alone (vendor-reported) — before a single campaign is sent. Get the order wrong and you are sending campaigns into a list that has never been given the chance to earn that number.

This piece is the setup, not the deep dive on any one flow. Each flow gets a sentence or two here and its own build guide elsewhere; what this page answers is the sequence, the segmentation that makes the sequence work, and the list-health settings that keep the whole thing sending.

What is an email marketing strategy for ecommerce, and why does it start with flows, not campaigns?

An email marketing strategy for ecommerce is the deliberate order in which you build automated flows, layer segments on top of the list those flows touch, and only then scale campaign volume into a list that flows and segments have already protected. It starts with flows because a flow triggers on a single subscriber’s own behaviour — they joined the list, they started a checkout, they bought — so it reaches the right person at the moment their intent is highest, with no send-day decision required from anyone on your team.

A campaign, by contrast, goes to a static or dynamic list on a day you choose, whether or not any individual recipient’s intent is high that day. Both matter. Neither is optional. But a campaign sent into a list with no flows underneath it is broadcasting into noise, and a flow built with no segmentation on top of it eventually sends the same message to a buyer and a browser who have nothing else in common.

Flows vs campaigns: which one earns the 41%, and which one earns the calendar?

Flows earn the 41% because they run on triggers, not on a publishing schedule — a welcome flow fires the moment someone joins, an abandoned checkout flow fires the moment someone leaves, and neither waits for a marketer to decide today is the day. Campaigns earn the calendar: launches, promotions, editorial content, and anything time-bound that no individual subscriber’s own behaviour would ever trigger on its own.

The mistake is not running both. The mistake is building campaigns first because they are visible — a send goes out, a report comes back — while flows sit half-built in the background, quietly leaving revenue unclaimed every day they are incomplete. Build the flow set first. A campaign calendar built on top of a working flow set reaches a list that is already warm; a campaign calendar built on top of no flows is the entire strategy, and it will plateau.

What order should a $3M–$30M brand build its core flow set in?

Build these five flows in sequence. Each flow depends on data the one before it makes reliable, and building out of order means debugging a flow’s targeting logic while a live flow underneath it is still misfiring.

Prerequisites: what has to exist before you build a single flow

Confirm three things before any flow goes live. First, Klaviyo’s Shopify integration is connected and the Placed Order, Started Checkout and Viewed Product metrics are actually populating — check a recent order or session against Klaviyo’s profile timeline, not just the integration’s connected status. Second, your sending domain is authenticated with SPF and DKIM records set at the domain level, not left on Klaviyo’s shared defaults. Third, someone owns list growth — a signup form or popup connected to a Klaviyo list — because a flow with no subscribers entering it cannot do anything, regardless of how well it is built.

Step 1: Turn on the welcome flow first

Trigger it on List Subscribed or Klaviyo Form Submission, never on Placed Order — a new subscriber who has not bought yet should hear from you before a campaign does, and a welcome flow triggered on the wrong metric either never fires for a non-buyer or fires for someone who already converted. This flow gets its own build guide; here it is one sentence: it is the first flow you build because every other flow assumes a subscriber has already been introduced to the brand.

Step 2: Build abandoned checkout before browse abandonment

Trigger on Started Checkout, add a flow filter excluding anyone who has Placed Order at least once since starting that checkout, and turn on Smart Sending so a subscriber does not receive this flow and a campaign on the same day. This is the highest-intent flow in the set — someone was one step from buying — which is why it is built second, ahead of colder-intent flows like browse abandonment.

Step 3: Build the post-purchase flow next, not last

Trigger on Placed Order and split immediately with a conditional split on Placed Order zero times before this order versus one or more times, so a first-time buyer and a repeat buyer never see the same post-purchase copy. Teams routinely build this flow last, after winback and browse abandonment, which is backwards — post-purchase is the flow that turns a first order into a second one, and every day it is missing is a day of first-time buyers with no path back.

Step 4: Layer browse abandonment once flows one through three are live

Trigger on Viewed Product and exclude anyone with an open Started Checkout flow, so a shopper who viewed a product and then started checking out does not receive two flows about the same item. This flow reaches warmer traffic than a cold subscriber but colder traffic than an abandoned checkout, which is exactly why it belongs fourth — it is real revenue, but it is not the revenue you leave on the table if flows one through three are not built yet.

Step 5: Build the sunset flow to protect deliverability

Define your own engaged-segment window on Opened Email or Clicked Email — the window itself is a judgement call your list’s send frequency and category should set, not a Klaviyo default — and move anyone who falls outside it into a flow that either re-engages them or suppresses them from future sends. Build this flow before you raise campaign volume, not after a raise has already dragged deliverability down. It gets a sentence here and its own build guide elsewhere; the point for this page is where it sits in the sequence, which is last among the core five and first among the flows most teams skip.

Segmentation is the lever most teams never pull

A flow reaches someone because of what they did. A segment decides what they see once they are inside a flow, or whether they see a campaign at all — and it is the difference between a list that behaves like one audience and a list that behaves like five. The segment worth building first splits the list on engagement: subscribers who opened or clicked inside your own defined window against everyone else. That single segment does two jobs — it protects deliverability by keeping campaign volume away from people who never open, and it lets a winback offer reach exactly the people it is meant for instead of the whole list.

A purchase-history segment is the second one worth building: zero orders, one order, two or more. Combined with the engagement segment, this turns a single campaign into four effective audiences without writing four separate campaigns — an engaged repeat buyer gets a different subject line judgement than an unengaged never-purchaser, even inside the same send. This gets a sentence here because the filter logic, the exact segment definitions, and how they interact with flow exclusion filters is its own build guide; the point for this page is that segmentation is not a nice-to-have layered on top of flows — it is the thing that keeps a campaign from undoing what the flows already built.

The step most teams get wrong

The step most $3M–$30M teams get wrong is not a missing flow. It is the order: they build browse abandonment, add a fourth and fifth campaign a week, and grow the list — all before a sunset flow exists to catch anyone who has stopped engaging. By the time someone notices open rates sliding, the unengaged share of the list has already been mailed dozens of times, mailbox providers have already downgraded the sending domain’s reputation with those providers, and the fix is now a deliverability recovery project instead of a five-minute segment build. Build the sunset flow — step five above — before you raise send frequency, not after deliverability has already told you it was overdue.

What list health and deliverability settings protect the whole programme?

Three settings do most of the protecting, and none of them is a one-time setup. Smart Sending — Klaviyo’s setting that suppresses a flow or campaign send to anyone who received another message from you inside a window you define — keeps a flow and a same-day campaign from landing in the same inbox twice. Domain authentication — SPF and DKIM configured at your sending domain, not left on a platform’s shared default — is what gives mailbox providers a consistent signal to build reputation against. And a defined non-engagement window, the input the sunset flow in step five runs on, is what keeps the list itself from becoming the deliverability problem: a list that never sunsets anyone accumulates addresses mailbox providers have stopped trusting long before your reports show it.

Smart Sending, domain authentication and a defined non-engagement window are settings to check on a schedule — is Smart Sending actually on, is the domain authentication still valid, is the sunset flow still catching people — because a setting configured once and never revisited is the most common way a working strategy quietly stops working.

How do you verify the strategy is working?

Verify at two levels, not one. At the flow level, check each flow’s own conversion rate and revenue per recipient inside Klaviyo’s flow analytics — a flow with a conversion rate near zero has a trigger or filter problem, not a copy problem, and no amount of subject-line testing fixes a flow that never fires correctly. At the programme level, check the split between flow revenue and campaign revenue over time: if flows are a shrinking share of total email revenue while your list is growing, campaigns are being sent to people your flows should have already reached, and the sequence in this guide is the thing to re-check, not the campaign copy.

Klaviyo’s 41%-of-revenue-from-flows figure is a benchmark, not a target — a blended average across 183,000 brands in different categories says nothing about what your own flow set should produce. Model what a fully built flow set could be worth for your own order volume and AOV with the flow revenue calculator rather than comparing your dashboard to someone else’s benchmark.

Who is this email marketing strategy not for?

This sequence assumes a Shopify or Shopify Plus store already running Klaviyo, or ready to be, with enough order volume for flow-level data to be statistically meaningful within weeks rather than months — in practice, the published floor: $3M or more in annual revenue. A brand doing a few hundred orders a month will wait a long time for any single flow’s conversion rate to be reliable, and this sequence is still correct for them, but the verification step will take longer to say anything useful. This is also not a guide for choosing between Klaviyo, Postscript or another platform — it assumes Klaviyo because that is what the filter names in this guide reference, and a different platform’s equivalent settings will carry different labels for the same logic.

A flow set built in this sequence still needs someone who owns it: checking Smart Sending is on, the sunset flow is still catching people, and campaign volume has not outrun what the segments underneath it can support — which is exactly the work our lifecycle flows engagements are built to run, from the welcome flow through the sunset flow and the segmentation that ties them together.