The short version: every brand needs the same six core flows, and four yes-or-no questions decide the rest. Is the product consumable? Is there a subscription? Is the purchase considered or fast? Is demand seasonal? Answer those and your required stack falls out mechanically, which is exactly how we scope builds for new accounts, and the economics justify the ruthlessness: flows generate nearly 41% of total email revenue from just 5.3% of sends, at nearly 18 times the revenue per recipient of campaigns, so whether the stack exists at all is the highest-leverage decision in the account. This post is the map: the universal core, the four questions, the full worked stack for each of seven verticals, and the 90-day build order, with every flow linked to its complete build standard.
The universal core: six flows every brand runs
No qualifying questions; if you sell DTC, these exist:
- Welcome, in its vertical variant (beauty, F&B, apparel, home, leather, outdoor), with the offer chosen per the offer framework. The consistent architecture across all seven: the segmentation question lives in the popup, the flow routes on it from email one, and the tail gates to engaged non-buyers.
- Abandoned checkout and abandoned cart, as two flows on two events, per the split architecture: Checkout Started gets the firm fast recovery, Added to Cart gets the gentle nudge, and the dedup rules keep one story per customer. The benchmark gap is the reason to build them properly: average recovery flows convert 3.33% of recipients while the top decile converts 7.69% at $28.89 per recipient against a $3.65 average, and the difference is architecture, not luck.
- Browse abandonment, repeat-view triggered, service-toned, the widest-reach flow in the account when its filters are honest.
- Post-purchase in its vertical variant (apparel's fit-check, beauty's skin cycle, F&B's first taste, considered purchases, outdoor's first trip), always as the two-system split: the owned transactional stack (order confirmations at 60 to 70% opens are the account's best-read real estate) and the delivered-triggered marketing arc.
- Winback, on thresholds from your own repeat cycle, never a borrowed 90 days, with the economics that make it always worth building: reactivation costs 5 to 10 times less than acquisition, and healthy programs win back 12 to 20% of the lapsed.
- Sunset, downstream of winback, on the dormancy math, because the unengaged tail taxes the deliverability of everything above it.
Question 1: is the product consumable?
Yes (supplements, beauty, F&B, CPG): add the replenishment flow (the variable-usage version for beauty and CPG), and it outranks winback in build order because it converts earlier, cheaper, at full margin: replenishment emails run 50 to 60% opens and 10 to 15% conversion with the highest click-to-open rate of any email type. Cross-sell takes the completion shape: stacks, routines, pantries. The whole consumable pipeline then runs in one fixed order: replenishment, then winback, then sunset, each flow catching the last one's misses.
No (durables): no replenishment; the second-purchase engine is category expansion instead (outfits, rooms), winback stretches to the occasion-driven build, and the care-content keepalive replaces the reorder reminder as the between-purchases thread.
Question 2: is there a subscription?
Yes: two flows become mandatory on day one. Dunning, because failed payments are the largest silent churn source and the recovery math is decisive: well-configured retries alone recover 45 to 70% of failed payments, with the three-email sequence front-loaded because the recovery curve is steepest in the first three days and decays sharply after two weeks. And the cancel-save flow branching on the cancellation reason, where the pause is the star: pause-before-cancel adoption is up 337% and roughly three of four pausers return. Subscriber suppression rules also now gate every other flow: subscribers never see replenishment, reorder pitches, or acquisition discounts, and their upcoming-order checkouts never trigger recovery.
No: skip both, but every consumable flow carries the subscription option in every send, because the replenishment flow's subscription bridge is how you build the program that eventually needs them.
Question 3: is the purchase considered or fast?
Considered (high AOV, long decisions: home, leather, jewelry, premium gear): recovery stretches to the four-email, five-to-seven-day build with logistics answers, financing framing (48% of consumers say BNPL availability influences their purchase decision), and the human reply-to branch above $1,000; browse abandonment gains weight relative to cart recovery because most intent never carts; post-purchase runs the delivery-wait arc that prevents cancellations during long lead times; and every flow obeys the category's manners: no countdowns, service concessions over discounts, patience as strategy.
Fast (consumables, low AOV): recovery compresses to 48 to 72 hours (F&B fastest of all), the mini or sampler becomes the recovery close (beauty's ramp rides 35%-plus sample-to-full-size conversion, F&B's variety pack does the same job), and free-shipping thresholds do the offer work at low AOVs where the shipping line is the real objection.
Question 4: is demand seasonal?
Yes (outdoor, gifting categories): every time-based threshold gains a season gate: winback enters at season start per the season-gated build, sunset stretches past a full cycle (the off-season silence that suppresses future buyers is the classic self-injury), post-purchase arcs wait for the opener per the first-trip logic, and the seasonal architecture governs the calendar. Gift-heavy brands add the gift/self split from the leather build, with the revenue concentration making the stakes plain: jewelry retailers earn 40 to 70% of annual revenue in Q4 alone.
No: day-counting works everywhere, which is the simple luxury evergreen brands rarely notice they have.
The worked stacks: all seven verticals through the tree
Supplements (consumable, usually subscription, fast, evergreen)
Core six + replenishment + stacking cross-sell + dunning + cancel-save, with FTC compliance riding every send and the goal property routing everything. Ten to eleven flows. Build emphasis: replenishment and dunning carry outsized revenue because the category is subscription-and-reorder economics wearing a storefront.
Beauty (consumable, sometimes subscription, fast, mostly evergreen)
Core six + variable-usage replenishment + routine cross-sell + the quiz feeding every flow's matched proof + the mini-size ramp in recovery. Nine to eleven flows. Build emphasis: the concern property and quiz data, because quiz-routed experiences convert at three to five times static baselines and every flow inherits the lift.
Food and beverage (consumable, often subscription, fastest, some seasonality)
Core six + replenishment on servings math + pantry cross-sell + dunning and cancel-save where subscribed, all on the fastest clocks in the library because 45% of category repeat purchases land inside 90 days. Build emphasis: the favorite question in post-purchase, the single data point the whole retention engine routes on.
Apparel (durable, no subscription, fast-to-moderate, seasonal)
Core six + back-in-stock (mandatory: the category's engagement king) + outfit cross-sell + season-gated winback. Nine flows. Build emphasis: the fit-check email, because size and fit drive 53 to 67% of apparel returns and the interception email attacks the category's biggest loss center directly.
Home and decor (durable, no subscription, considered, mildly seasonal)
Core six with the considered variants + room cross-sell + occasion winback + the swatch-request branch. Eight to nine flows. Build emphasis: patience mechanics: the category opens and clicks above average while converting below it, so the flows that respect the long cycle win it.
Leather goods and jewelry (durable, no subscription, considered, gift-seasonal)
Core six with considered variants + the gift/self split + occasion winback + category-expansion cross-sell + personalization mechanics (77% will pay more for customized accessories, and customization correlates with 40% fewer returns). Eight to nine flows.
Fishing and outdoor (durable, no subscription, moderate, intensely seasonal)
Core six, all season-gated + gear-system cross-sell + back-in-stock + first-trip post-purchase + the care-ritual keepalive. Nine flows. Build emphasis: the species property crossed with the calendar, the personalization engine of the entire vertical.
The conditional add-ons, regardless of answers
- Back-in-stock, if anything ever sells out: the highest per-recipient performer in the account (7.9% email conversion, 8.5% SMS in Klaviyo's data), and the stakes of not having it are the switching data: 45% of shoppers buy from a different retailer when they hit a stockout and 62% have switched brands over one. Mandatory for drop and scarcity brands; pointless only for never-out-of-stock catalogs.
- Cross-sell, once there is a second thing to genuinely recommend: gated on first-product belief in every variant, one pairing at a time, judged partly on the return rate of cross-sold items.
- SMS lanes on the moments that earn them, per the channel division: back-in-stock and delivery-day first, one checkout nudge second, nothing else until those prove out, because SMS flows earn 45.2% of SMS revenue from 7.6% of sends precisely when they stay confined to time-critical moments.
The 90-day build roadmap from zero
Days 1 to 15: the money layer. Welcome (full vertical build, popup question included) and abandoned checkout, the two highest-revenue-per-build-hour flows in every account, plus the transactional stack at least styled. This fortnight typically captures the majority of the total build's quick-win revenue.
Days 16 to 40: the category heavy-hitters. Replenishment for consumables or back-in-stock for everyone else (both, where both apply), the cart flow and its dedup rules, and dunning plus cancel-save immediately if subscriptions exist. These are the flows audits most often find missing and most often find worth the most.
Days 41 to 70: the lifecycle layer. Post-purchase's full two-system build with its vertical centerpiece (the fit-check, the favorite question, the setup check), browse abandonment with honest filters, and cross-sell's first pairing map.
Days 71 to 90: the hygiene layer. Winback on real repeat-cycle math, sunset with its threshold from the dormancy guide, the SMS lanes that earned their place, and the first pass of the flow audit run against your own fresh build, because the builder's audit catches what the builder missed.
The five most common stack mistakes
- One recovery flow doing two jobs: a single flow on one event mis-times both the cart and checkout audiences; the split exists for a reason.
- Winback before replenishment in consumable accounts: building the safety net before the primary system, backwards by both conversion rate and margin.
- Flat calendar math in seasonal brands: the 90-day winback and open-based sunset quietly suppress next season's buyers mid-cycle.
- The unbuilt back-in-stock flow at a brand that sells out monthly: the account's best per-recipient economics, left as an app-setting checkbox.
- Flows without owned metrics: a stack nobody measures decays silently, which is what the quarterly cleanup exists to catch.
Frequently asked questions
How many Klaviyo flows does a brand actually need?
Six universal (welcome, checkout, cart, browse, post-purchase, winback, plus sunset as hygiene), then two to four conditionals decided by consumability, subscription, purchase speed, and seasonality. Most brands land between eight and twelve live flows; the seven vertical stacks above show the exact counts.
Which flow should be built first?
Welcome and abandoned checkout: highest revenue per build-hour in every account we open. The 90-day roadmap sequences the rest: money layer, category heavy-hitters, lifecycle, hygiene.
What is the most commonly missing high-value flow?
Replenishment for consumables and back-in-stock for everyone else. Both routinely top the recovered-revenue list in audits precisely because they are the least-built, and both carry the strongest published engagement numbers in their categories.
Do subscription brands need different flows?
Two additions (dunning, cancel-save) and one rule change: subscribers are suppressed from every acquisition-flavored send. The dunning math alone justifies the priority: retries recover 45 to 70% of failed payments before an email is even written.
How do I audit my existing stack against this?
Run the 15-point flow audit: coverage first, then events, timing, suppression collisions, offers, and measurement, in the fix order that recovers revenue fastest, then keep it healthy with the quarterly cleanup.
Sources
- Klaviyo. Email marketing benchmarks.
- Klaviyo. Abandoned cart benchmarks.
- Klaviyo. Back-in-stock notification data.
- Klaviyo. SMS marketing benchmarks.
- Eightx. Win-back and reactivation benchmarks.
- Eightx. Apparel returns and fit data.
- SmartMail. Replenishment email benchmarks.
- Baremetrics. Failed payment recovery data.
- Digital Applied. Dunning recovery-curve data.
- Recurly. Cancellation flow and pause data.
- Doss. Stockout switching research.
- CUFinder. Food and beverage benchmarks.
- Justuno. DTC furniture and BNPL data.
- Free Yourself. Sampling conversion data.
- Visual Quiz Builder. Quiz conversion benchmarks.
- Kickflip. Product customization statistics.
- Branvas. Jewelry seasonal revenue statistics.
- Klaviyo. Home furnishings engagement patterns.