Klaviyo flow templates

Six flows, and the chart Klaviyo publishes

The core set for a store is a welcome series, an abandoned cart, a browse abandonment, a post-purchase flow, a winback, and a sunset. Klaviyo’s getting-started guide names welcome, abandoned cart, post-purchase, and winback as the flows to learn first. Browse abandonment and sunset keep those four from doing the wrong job. Browse catches a product view that never reached checkout. Sunset is the last mail to someone who has stopped responding, before you suppress them.

Klaviyo’s article on flow sending frequency says to start from the pre-built templates in the flow library. It publishes one chart of the frequencies it has found work for the average brand. The chart is a grid of marks. It does not match every email count in the individual flow articles. Read the chart as the library’s default marks. Read the flow article when you need the email count and the filter. Where they disagree, this page says so. Klaviyo also says the right gap varies by business, and that you can test two delays with a conditional split or a split on a random sample. Keep the content the same, so the delay is the thing that changed.

• Welcome series: immediately, and again at 4 days.
• Abandoned cart: 4 hours, and again at 1 day.
• Browse abandonment: 2 hours.
• Post-purchase: immediately, 3 days, and 1 week.
• Winback: 2 months, and again at 3 months.
• Sunset: 1 day, 1 week, and 2 weeks.

Welcome series

The welcome series is list-triggered. It starts when someone is added to the list on the trigger. It does not start because they viewed a page. The form, the subscribe page, or the API has to write to that same list. A form that subscribes people to one list while the flow listens to another will sit there marked live and send nothing.

Klaviyo’s welcome-series article describes a pre-built flow of three emails. The first has no delay. It goes immediately after the person is on the list. With double opt-in, which is the default for every list, immediately means after they confirm, not after the form is submitted. The second email is 3 days after signup. The third is 4 days after the second, about a week after signup. The article calls that the cadence for a standard series, and says you can change what each email is for. The pre-built goals are an introduction and a preference ask, then a social ask, then best sellers. Replace a social ask that points at an account you do not maintain.

The frequency chart marks only two welcome sends: immediately, and at 4 days. That is fewer emails, and a different second gap, than the three-email pre-built. Open the flow and read the delays on the canvas. If the account used the pre-built, expect three messages and a 3-day gap before the second. If someone shortened it to the chart, expect two. A multi-day delay in front of the first email makes a promised code late. The first message is also the one email Smart Sending is most likely to skip. Email Smart Sending defaults to 16 hours, and a skipped message is not retried when the window ends.

List-triggered flows do not re-enter by default. Someone who unsubscribes and later resubscribes does not go through the series again unless you change that setting. Each email is still draft, manual, or live on its own. A draft first email is skipped, and a later live email can still send.

Abandoned cart

An abandoned cart is not a separate Klaviyo event. On a checkout flow, the trigger is the checkout metric. Klaviyo’s abandoned-cart article calls it Started Checkout. The Shopify data reference calls the same family of events Checkout Started. Older help text uses both names. Open Analytics, then Metrics, and build the flow on the checkout metric the account is receiving.

The filter that makes the flow an abandonment flow is Placed Order zero times since starting this flow. Profile filters are checked at entry and again before each message. People who buy are skipped, and they stay in the flow so the next step can check again. A large “fails profile filters” count is the usual sign that the filter is working. Placed Order and Ordered Product are different events and sync seconds apart. A filter pointed at the wrong one skips the people you meant to email, or emails the people who already bought.

Klaviyo’s pre-built abandoned cart waits 4 hours before the first message. The same article recommends a first message about 2 to 4 hours after checkout starts, and a second 20 to 48 hours later. The frequency chart marks those two sends at 4 hours and at 1 day. The article also says research across Klaviyo stores found that 2 or 3 abandoned-cart messages perform best. Hold a coupon for the last email, or for someone who has never purchased. A coupon in the first email trains people to abandon. If the integration syncs slowly, a short delay can send before the order arrives. Klaviyo’s floor for a short delay is 1 hour and 15 minutes.

The checkout URL on a Shopify event is {{ event.extra.checkout_url }}. The repeating item block is dynamic and reads that event. If an Added to Cart flow is also live, it needs its own skip for people who started checkout, or one person gets two series. On Shopify, Added to Cart requires Track behavioral events. A checkout-metric flow does not, because the event comes from checkout.

Browse abandonment

Browse abandonment triggers on Viewed Product. The person did not add anything and did not start checkout. Klaviyo’s browse article says that is a weaker signal than a cart, and that the message should be a lighter touch than the abandoned-cart mail. Use a pre-built flow when the integration has one, because the product tags are easy to break if you rebuild them.

When you build it yourself, the profile filters in that article are Placed Order zero times since starting this flow, Started Checkout zero times since starting this flow, and has not been in this flow in the last 30 days. Add Added to Cart zero times since starting this flow when an Added to Cart flow is also live. Without that filter, a person who views a product and then adds it can sit in both series.

The delay in a flow you build should be at least one hour. The frequency chart marks a single browse send at 2 hours. One email is the chart. The article allows more delays and more messages if you add them. The cart block does not belong in this email. The cart block is dynamic and repeats per line item. The browse block is static, and switching the block type erases the tags. Klaviyo’s tags for the viewed product are event.ImageURL for the image, {{ event.Name }} for the name, {{ event.Price }} for the price, and {{ event.URL }} for the product page. If the person viewed several products, “has not been in this flow” shows the first item. A filter on viewed product since starting the flow shows the most recent.

On Shopify, Viewed Product is an onsite event. It needs the app embed. In the EU, EEA, UK, and Switzerland it follows the store’s customer privacy settings, so a browse flow misses visitors who have not consented. An empty Viewed Product metric is a tracking problem. Rewriting the email will not create the event.

Post-purchase

A post-purchase flow triggers on Placed Order. Klaviyo groups a thank-you, a cross-sell, an upsell, and a review request under that name. The setup article is specific about delays. A thank-you should not wait. A review, a cross-sell, or an upsell should. The frequency chart marks three sends: immediately, 3 days, and 1 week.

Order and shipping confirmations are a different flow. Klaviyo says a transactional message cannot contain a marketing effort, and that those confirmations should stay in their own flows so every buyer gets them. The example in the post-purchase article, for a brand that ships in about 9 days, is a thank-you, an optional instructions email timed to that shipping window, a review request, and a cross-sell. That is a longer marketing series than the three marks on the chart. Keep the confirmation out of the marketing series, then decide whether this account’s marketing flow is the three-mark chart or the longer example.

A review that fires on Placed Order, before the package arrives, asks someone to review a product they do not have. Time that message to delivery, or trigger it from a fulfillment metric, when the catalog ships slowly. The 3-day mark is the average-brand chart, not a delivery date. A conditional split on Placed Order equals 1 over all time separates a first order from a later one. The first-order path can point at the list or a loyalty program. The repeat path should speak to someone who has already bought.

Smart Sending matters here because a new buyer can be inside the welcome series and the post-purchase flow in the same week. Email Smart Sending defaults to 16 hours. A skip is not sent later. SMS and push use a separate 24-hour window. Check the setting on each message. Some pre-built flows turn the skip off on a particular email.

Winback

A winback flow also triggers on Placed Order, with the profile filter Placed Order zero times since starting this flow. The first delay is the product. Klaviyo’s winback article says to set it a little longer than the store’s average time between orders, so you write to someone who has missed a purchase. A couch and a bottle of shampoo do not share that number. If predictive analytics is available, the method in the article is a segment of people with at least 2 Placed Order events in the last 2 years, exported, then averaged on the Average Time Between Orders column. If it is not available, answer when a customer would need the product again, and add a little time beyond that.

The winback article recommends three emails. The first is the short one, with three to six products and, if you use one, an incentive. The second reminds them of that incentive and puts a date on it. The third is the last call, with a visible unsubscribe. The frequency chart marks two sends, at 2 months and at 3 months. Those marks are the average-brand default in the library. They are two emails, where the article recommends three, and 2 months is not this catalog’s buying cycle until you have checked the orders.

A product people repurchase on a schedule belongs in a replenishment flow, not in this one. If the store also sells through Buy with Prime, Klaviyo says to run a second winback on that order metric, and to add Placed Order (Buy with Prime) zero times since starting this flow on the ecommerce winback. Otherwise a Buy with Prime reorder looks like a lapse.

Sunset

A sunset flow is the last mail before you stop emailing someone who does not respond. It triggers when a person is added to a sunset segment. Klaviyo’s sunset article gives the definition that matches the prebuilt Sunset (Email) segment. The person can receive email marketing. The profile was created at least 180 days ago. They have Received Email at least 5 times in the last 72 weeks. Over all time they have Opened Email zero times, Clicked Email zero times, been Active on Site zero times, Viewed Product zero times, Checkout Started zero times, and Placed Order zero times. Suppression recommendations elsewhere in the product use a stricter created-at of 365 days. Build the segment you actually intend to suppress, and read the definition before you send.

The flow should contain 1 to 3 emails. The article says not more than 3, because mailing unengaged people repeatedly hurts delivery. The frequency chart marks three sends, at 1 day, 1 week, and 2 weeks. After the last email, add a grace period so a late open or click still counts. The article uses 3 days when you send daily, and 7 to 10 days when you send less often. Profile filters on the trigger should require Opened Email zero times since starting this flow and Clicked Email zero times since starting this flow. Someone who opens or clicks fails those filters and skips the rest, including the property update.

The property update sets a Boolean named Unengaged to true. Klaviyo is explicit that this does not unsubscribe anyone and does not suppress them. You suppress later, from a segment where Unengaged is true, with Suppress current members. That action covers the members in the segment at that moment. People who qualify afterward are not included, so the suppress has to be repeated before a later campaign. There is no native flow action that suppresses email. Email suppression also does not stop SMS or push.

Profiles already in the segment when you create it do not enter the flow. The flow triggers when someone newly qualifies. Adding past profiles queues the existing members, and only into actions that are already manual or live. Add them at the beginning if they should see the first email. Throughout the flow places them by when they joined the segment, which can skip the earlier emails. Apple Mail Privacy Protection can record an open the person did not make. A segment that requires zero opens over all time will exclude those profiles. Klaviyo’s winback article warns about the same inflated opens when a flow depends on an open.

Read the delay that is on the canvas

A template is not the flow in the account. For each of the six, read the trigger metric, list, or segment, the profile filter, the delay, the number of messages that are live rather than draft, and a preview against a real event. The chart’s 2-hour browse email and the welcome article’s three-email week are both published by Klaviyo. Only one of them is what this account will send tomorrow. The chart also marks an anniversary send at 1 year. That is a separate template, with one email on the anniversary, and it is not a substitute for winback.

KlavyFlow is a read-only view of the connected Klaviyo account. The flow timeline places each message on the elapsed day, so a 4-hour cart email and a next-day follow-up sit next to the creative, and a winback that waits two months sits on that later day. It does not change a filter, turn a draft live, or send mail. Performance for one flow is a separate view over 7, 30, or 90 complete account-local days. A missing result stays missing. Change the series in Klaviyo. Use the timeline to see which of these templates is actually on, in each client workspace.