Short answer
Start with welcome and abandoned checkout, then add browsing, post-purchase and win-back flows. Rank by intent, eligible volume and measured contribution.
The answer in detail
Launch welcome and abandoned-checkout flows first, then browse abandonment, post-purchase and win-back. Use revenue per recipient to establish a starting order, then adjust for how many eligible people actually enter each flow. A flow with excellent theoretical economics and no working trigger earns nothing.
The benchmark
The original article cites Klaviyo's October 2, 2026 benchmark report covering more than 110,000 accounts. Its email averages are attributed revenue in US dollars, not incremental profit or a forecast for your store.
| Flow | Typical trigger | Average revenue per recipient |
|---|---|---|
| Abandoned checkout | Checkout started without purchase | $6.77 |
| Welcome | List signup | $5.75 |
| Browse abandonment | Product viewed without cart addition | $2.08 |
| Post-purchase | Order placed | $1.80 |
| Win-back | Extended absence of purchases | $0.82 |
Volume can reverse the ranking. A women's fashion brand described by the contributors had 219 abandoned-checkout recipients and 1,240 browse-abandonment recipients over 90 days. Multiplying those counts by the benchmark averages produces about $1,500 and $2,600 respectively. These are illustrative projections, not the brand's measured results.
eligible entries × reference revenue per recipient = planning estimate of attributed revenue
Use one observation window and distinguish entry counts from message deliveries. A three-email sequence does not necessarily have three times as many recipients as a one-email sequence.
The rule
Rank by buying intent, then by eligible volume. A recent checkout or signup gives you a clearer reason to contact someone than a long-past purchase. Klaviyo's getting-started guide also prioritizes welcome and checkout recovery.
Finish the operational setup before adding another flow: events arrive, all intended messages are live, purchase exits work and recipients meet the appropriate marketing rules. An anonymized fashion brand in the source had 45 flows, 25 still in draft. Its first welcome message accounted for 97% of welcome revenue because the next two had never been activated.
Permission is a separate gate from event tracking. A known email address or a checkout event does not automatically authorize marketing in every jurisdiction. EU data-protection and electronic-marketing rules, UK PECR and US CAN-SPAM have different scopes and requirements. Keep a defensible, market-specific eligibility policy. GDPR consent conditions do not themselves replace the applicable electronic-marketing rules.
The steps
- Inventory every flow, its live/draft/manual status and 90-day entries. Investigate zero counts. At a small textile-homewares business in the source, 301 accounts had been created without their addresses reaching the intended list.
- Build the welcome flow around the list-subscription event. Deliver what the form promised. See collecting emails without a discount and welcome-email content.
- Add checkout recovery. The source's Klaviyo setup uses Started Checkout, a first message after two to four hours, a second 20–48 hours later, and an exit filter for people who purchase. Treat those timings as a starting configuration. Apply the appropriate permission and suppression filters. The textile business reported 9.5% unsubscribes over 30 days in a flow without its intended marketing-consent filter.
- Add browse abandonment, triggered by Viewed Product after at least an hour in the source setup. Exclude people who started checkout, purchased or recently entered the flow; its reference interval is 30 days. Address real purchase barriers, rather than merely repeating the product image.
- Add practical usage information after delivery, review requests and a replenishment or second-purchase sequence. Only then build win-back around the actual repurchase cycle. Separate service messages from marketing.
- Review monthly using revenue per 1,000 delivered emails, alongside unsubscribes, complaints and contribution. Keep the working version before changing a flow. Use a holdout to test incrementality rather than equating attribution with causation.
The pitfall
A large dormant list feels urgent, but win-back is the weakest benchmark per recipient. Building it first can leave today's new subscribers and checkout abandoners without any follow-up.
An outdoor brand in the source had flows contributing only 8% of email revenue in fall 2026. A language-dependent signup-window problem limited welcome entries to about 500 monthly. Reconnecting checkout recovery reportedly produced about €15,000 in attributed September revenue. The source's practitioner estimate that healthy accounts often generate 25–40% from flows is unpublished field experience, not an industry standard.
What to tell your team
“In this fictional month, 1,200 people start checkout without buying and 900 subscribe, with no automated follow-up. We will launch those two eligible flows, verify their triggers and exits, then compare revenue per 1,000 emails and incremental contribution.”
