Most Klaviyo accounts I audit look busy from the dashboard and thin from the inbox. Twenty-six flows turned on, eight of them stepping on each other, and the operator has not read the actual send log in months. The numbers Klaviyo surfaces at the top of the account say everything is fine. The inbox says otherwise.
This playbook is the version of Klaviyo lifecycle marketing I wish someone had handed me the first time I owned a DTC retention program. Eleven flows, how they fit together, what each one is actually for, and where the data behind them comes from. It is the map I use now when a new client opens their Klaviyo and asks where to start.
What this playbook covers and what it does not
The playbook is built around one assumption: you already have a working Shopify store, a Klaviyo account, and enough traffic that lifecycle email can move real dollars. If you are pre-launch or pre-product-market-fit, none of this matters yet. Go find your first hundred buyers a different way.
Everything here is platform-biased toward Klaviyo plus Shopify because that is the stack I have rebuilt most often. The underlying logic transfers to Omnisend, Attentive, or a homegrown stack running on Postmark and a Postgres segmentation layer. The flow names change. The decisions behind them do not.
I am not going to tell you that lifecycle email will save a brand that does not have a real product. It will not. What it will do is stop quiet margin bleed. A weak welcome series costs a brand roughly one first-purchase customer in every ten opt-ins it could have converted. A missing win-back costs three to five percent of last year's purchasers going dormant every quarter. The math adds up into real money over a year.
The eleven flows, ranked by how much they move revenue
If you want the short version: ship the first four, measure for ninety days, then ship the rest. Most brands I work with never need flows twelve through twenty. Twenty flows is a vanity metric.
- Welcome series (opt-in, no purchase yet). The first-impression flow. Gets the brand into the inbox, sets delivery reputation, and converts the coldest buyer you will ever have. Welcome series architecture for DTC walks the specific sequence I ship.
- Abandoned checkout. The one flow every store already has and usually has tuned correctly by accident. If your flow revenue is mostly coming from this one, it is because nothing else is working, not because this flow is special.
- Post-purchase. The most underbuilt flow in DTC. Everyone ships a "thanks for your order" email and calls it done. The real play is a seven-touch sequence that builds product understanding, sets up the review ask, and seeds the next purchase. The post-purchase flow that adds LTV is the exact sequence I use.
- Browse abandonment. Lower revenue per send than cart abandonment, higher total reach. Whether this deserves the spend on your catalog depends on pattern, not on best-practice. Cart abandon vs browse abandon is the decision framework I use with clients.
- Win-back. For buyers who have gone dormant past your typical repurchase interval. The cadence question here is harder than most operators think. Win-back email cadence for DTC has the field-note version of how I answer it.
- Back-in-stock. Triggered, not scheduled. Revenue per send is usually the highest in the account because intent is already locked in. The only flow that should have more aggressive SMS companion logic than email logic.
- Price-drop. For subscribers watching a specific SKU. If you have a wishlist feature, this is the flow that makes the wishlist worth building.
- Replenishment. For consumables. Fires at the predicted end-of-supply date, not on a fixed cadence. Needs clean order data in Klaviyo, which most stores do not have. The Klaviyo and Shopify sync gaps is what you read before you try to build this one.
- VIP thank-you. For buyers who cross a revenue threshold you set. Light-touch. This is brand work, not conversion work, and you have to measure it differently.
- Review request. Usually attached to the post-purchase flow rather than standalone, but worth naming separately because it lives on its own trigger (order fulfilled plus product-specific delay).
- Transactional companion. The forgotten one. Order confirmation, shipping confirmation, delivery confirmation, refund confirmation. These are deliverability currency, not marketing. Transactional email deliverability covers the failure mode that cost a client I was working with three months of lost reputation.
The three decisions that determine whether any of this works
1. Segmentation, not just triggers
Klaviyo flows can technically run without segmentation. Every opt-in enters the welcome series, every cart abandon gets the same three emails, every purchaser sees the same post-purchase sequence. That is the default configuration and it is why most Klaviyo accounts underperform.
The segmentation decision that moves the most revenue is the cheapest one to make: first-time buyer versus repeat buyer. Your welcome series, your post-purchase flow, and your win-back all need this split because the message, the offer, and the timing are different for each. Klaviyo segmentation patterns for small-catalog DTC is the pattern library I pull from when setting these up.
For small catalogs (under fifty SKUs), you do not need product-level segmentation. Category level is enough. For large catalogs, you need individual-SKU behavior and Klaviyo's default data model starts to strain. That is a rebuild conversation, not a flow-configuration one.
2. SMS is not an add-on, it is a reframe
Most DTC operators treat SMS as an extra channel bolted onto email. Send the email, send an SMS version of the same thing, hope for the best. This is how you end up annoying the buyers who opted into both.
The right question is: for each flow step, which channel carries the message, and does the other channel even need to fire. Abandoned checkout email one hour in, SMS at four hours only if no purchase yet, email at twenty-four hours. That sequence beats duplicate blasts on both channels at every interval. SMS vs email decision matrix for DTC is the decision log I use to make these calls with clients.
3. Deliverability is the floor
You can have the best sequence in Klaviyo and still lose eighty percent of your sends to spam because the domain is not authenticated, the warm-up was done wrong, or a single badly-targeted campaign tanked the reputation three months ago. No amount of flow optimization fixes a deliverability problem.
DMARC, SPF, and DKIM are the baseline. You need all three, correctly configured, before the domain will reliably land in primary inboxes. Email deliverability: DMARC, SPF, and DKIM plainly is the walkthrough I send to operators who need this done fast.
“A weak welcome series costs a brand roughly one first-purchase customer in every ten opt-ins it could have converted. That is the quiet bleed.
”
Where Klaviyo's data model helps and where it lies
Klaviyo sells "predictive" metrics. Predicted CLV, predicted next order date, expected date of next purchase. These are useful if you understand what they are and dangerous if you do not.
The predicted CLV number in Klaviyo is computed from purchase history within Klaviyo only. If a buyer has two orders in Klaviyo and fifteen in a subscription system that never syncs to Klaviyo, the predictive number is calibrated against the two. I have seen accounts where the "top CLV" segment was roughly the population of buyers with exactly two orders, because that was the shape of the data Klaviyo had, not the shape of the customer base.
Predictive LTV in Klaviyo: how much to trust the number is the contrarian essay I ended up writing after the third client asked me why the predicted numbers did not match their accounting.
The second place Klaviyo's data model strains is on the Shopify sync. Klaviyo syncs product data, order data, and customer data from Shopify through a documented but quietly brittle pipeline. Custom line-item properties, subscription data, and tax-inclusive pricing all have edge cases. If your flows depend on any of these fields (replenishment flows especially), you need to verify the sync is doing what you think it is. The field notes on the Klaviyo and Shopify sync gaps covers what I have found in production.
The calendar layer: campaigns versus flows
Flows are evergreen. Campaigns are timed. A working lifecycle program has both.
The campaign layer is easy to overbuild. Every brand has the urge to ship a campaign every Tuesday because that is what "email marketing" looks like on LinkedIn. Most of those campaigns hurt deliverability more than they help revenue. The right cadence for a small DTC brand is usually one campaign per week, sometimes two, and the content has to be good enough that nobody would unsubscribe if you described it to them in person.
Black Friday is the exception. For most DTC brands, Q4 campaigns drive thirty to forty percent of full-year email revenue and the deliverability and planning work has to start in August. Planning a Black Friday email calendar that does not implode is the decision log I use to structure the calendar without cannibalizing other sends.
Where this fits into a full DTC stack audit
Lifecycle email does not live in isolation. It depends on clean attribution, clean product data, clean customer identity, and clean event tracking. If any of those are broken, your flows will fire on the wrong trigger or with the wrong data, and every optimization on top will be noise.
This is why the first thing I do with a new retention client is an audit of the full stack: Klaviyo, Shopify, attribution, CAPI, consent, transactional. The DTC stack audit I productized was built exactly for this situation. It is the shortest path from "I think my flows might be broken" to a written map of what is actually wrong and what to fix first.
If you want context on the tracking layer that sits underneath Klaviyo flows, the CAPI field guide is the cross-cluster companion piece. A broken CAPI setup can make your Klaviyo attribution numbers look wrong even when Klaviyo itself is fine. And on the Shopify side, the Shopify infrastructure architecture piece covers the theme and backend choices that determine whether Klaviyo's sync has good data to work with.
FAQ
How many flows should a small DTC brand actually run?
Four to start. Welcome, abandoned checkout, post-purchase, and browse abandonment. That covers most of the revenue most lifecycle programs produce. Adding more flows before those four are measured and working is the most common way I see operators dilute their own program.
Does Klaviyo replace an email marketing agency?
For most brands under $5M in annual revenue, yes. The question is usually whether you have someone internally who can own the program or whether you need a fractional partner to handle it. Flows are simple enough that an in-house operator can run them once the architecture is in place.
Is predictive CLV in Klaviyo reliable?
It is directionally useful and numerically unreliable. Treat it as a segmentation signal ("which buyers are higher value") and not as a forecasting tool ("what will this customer spend"). The underlying model only sees purchases that sync into Klaviyo, which is often a subset of the real history.
When do I actually need SMS?
When email alone is not enough to reach your buyer at the moment the flow step needs to fire. Abandoned checkout at the four-hour mark is a strong SMS use case. Post-purchase thank-you is not. The decision is per flow step, not per brand.
What is the single biggest mistake you see in Klaviyo accounts?
Turning on every flow the platform suggests without a clear model of why each one exists. Twenty-six flows turned on, half of them stepping on each other, is worse than eight flows that each earn their spot. Fewer flows, better built.
What to read next in this cluster
If you are starting from zero and want the architecture: the welcome series piece is the entry point. If you have flows running and want to know which to prioritize rebuilding: the post-purchase walkthrough and the segmentation patterns cover the two highest-leverage rebuilds. If you are dealing with deliverability problems right now: the transactional deliverability postmortem is the fastest path to diagnosis.
And if you would rather have someone look at the whole thing and tell you exactly what to fix, the DTC stack audit is built for that conversation.