A browse abandonment email is a message sent to a shopper who looked and left.
Every platform documented here agrees on that. They disagree on what counts as looking.
One sends it the moment someone opens a product page. Another sends it only to people who never reached a product page, and sells the product-page version under a second name. A third calls it something else again.
Same two words. Different shopper.
The browse abandonment email, as the platforms document it. Two run a two-step ladder for pre-cart browsing and put the same label on opposite rungs. Four more reuse a third name. The published wait before the first message runs from one hour to twenty-four, with the stated reasons opposed. And the rates that circulate for it are credited to a report that publishes four other message types instead.
What is a browse abandonment email?
It is an automated email sent to someone who viewed something in an online store and left without buying it.
Klaviyo’s help centre draws the line at the cart. “In browse abandonment flows, a visitor doesn’t have to add an item to their cart to trigger this flow. All a site visitor has to do is view an item and move on.”
Braze describes its template the same way, for users “who have browsed products but did not add them to their cart or place an order.”
How is it different from an abandoned cart email?
The cart email follows someone who picked an item up. This one follows someone who only looked, so the signal behind it is weaker.
The first vendor says so. Visiting a product page “doesn’t indicate the same level of interest as adding an item to a shopping cart and beginning the checkout process”, so its advice is to make these messages a “lighter touchpoint” than the cart flow.
A guide from Omnisend ranks the stages the same way. Browse abandonment is marked “Low” intent, product abandonment “Medium”, cart abandonment “High”.
Do those two words mean the same thing everywhere?
No. Two of these platforms use the same two words for two different shoppers, and each sells a second product for the other’s shopper.
On the first, the flow fires on a product view. Its documented trigger is the Viewed Product
metric.
On the second, the same name means the opposite. Its Browse Abandonment automation re-engages contacts who visited ”– browsing a homepage, category page, or sale page – but left without viewing a specific product or adding anything to their cart.”
Which shopper does “browse abandonment” actually mean?
It depends whose documentation is open. One means the product page, the other every page before it.
The first covers the pre-product visit with a separate, earlier flow. It is “the first step in the customer journey, sending before even a browse abandonment flow”, and it “includes anyone who visited your website, but didn’t look at or buy any products.”
The second covers the product view with an automation called Product Abandonment, for contacts “who viewed a specific product page in your store but left without adding it to their cart.”
Two behaviours, two products, the labels swapped.
What else is this message called?
Depending on the platform it is an abandoned browse, a product abandonment, a product retargeting email, or an abandoned product browse.
Mailchimp does not call it browse abandonment or abandoned browse. Its version invites shoppers back “with a product retargeting email”, and it is “a classic automations feature, which is only available to paid accounts that have previously created a classic automation.”
Seguno uses the abandoned browse name for its own automation and “Abandoned product browse” for the Shopify template it sits on.
What event actually starts it?
On several of these platforms a single product view is enough, and one of them requires that tracking to be installed first.
The Viewed Product metric is the stated trigger on one. “This flow thus requires Viewed Product
tracking to be enabled on your website.”
A second enters users “when a user performs the ecommerce.product_viewed event”. The product-side automation elsewhere adds a stock condition, “with a default filter that requires the product to be in stock.”
Does viewing more than one product change anything?
On one platform it does. Bloomreach’s use case will not send until the shopper has seen at least two products.
Its conditions are that the customer “was not part of any Abandoned browse campaign in the last 7 days, hasn’t purchased anything in the last 7 days and has seen at least 2 products on the website.”
Elsewhere here, one view triggers it.
Does a category page count as browsing?
On one platform it is the entire point of the flow. On another it belongs to a different automation.
The page-view trigger is “preset to the Viewed page . By default, it fires for all page views.” It warns against pointing it at products, because “Product Abandonment already covers product page views.”
The other builds the exclusion into filters. Its site flow triggers on
“Active on site” and screens out anyone with Viewed Product since the flow started.
What happens when someone adds the item to their cart?
Normally they leave this flow and enter the cart one. On two platforms that hand-off is something you build yourself.
Braze exits users who “Performs ecommerce.cart_updated , ecommerce.checkout_started , or ecommerce.order_placed”.
One vendor documents a chain. “In Browse Abandonment , add Viewed product as an exit condition.” Then “In Product Abandonment , add Added to cart as an exit condition.”
Klaviyo makes its version conditional. The Added to Cart filter “Only applies if you also have a
flow triggered by the Added to Cart metric, as otherwise, customers could get emails from both
flows”.
Does every platform stop at the cart?
No. Dotdigital’s definition runs to the order instead, so a shopper who filled a cart is still inside it.
Its program sends reminders “after they view products, but don’t complete a purchase.” Each of its three triggering routes ends the same way, with contacts who “browse items, and then leave the store without completing an order.”
A wider group than the other definitions describe.
Who is allowed to receive one?
Only a visitor the platform already recognises. An anonymous browser cannot be sent one at all.
One vendor states it outright. The automation “only fires for identified contacts – visitors whose email address is linked to a tracking cookie in their browser.” The cookie is placed when a contact signs up through one of its forms, clicks a link in one of its emails, or submits an email at checkout.
Then the consequence. Visitors who have done none of those things “cannot be tracked, and the automation will not trigger for them” even where the address is already on the list.
Why do so few visitors qualify?
Because a visitor is not identified until they click a platform email, submit a form, log in, or enter an email at checkout.
The same vendor titles its own FAQ entry “Why isn’t Browse Abandonment triggering even though I have thousands of contacts?”
The first vendor tracks only “identifiable browsers” — people who have engaged, submitted a form, or been “cookied”. Another requires that “Email is not blank”, and Shopify’s automations “send only to customers subscribed to email marketing by default.”
Is one route in stricter than the others?
Yes. On one platform the only way in is a click from that platform’s own email campaign.
“To receive a product retargeting email, a subscriber must first navigate to your store via a campaign link. Cookies will begin to track their activity for up to one year—or until they clear their cookies.”
Dotdigital accepts three — a click from one of its campaigns, a visitor who submits an email address or mobile number, or a logged-in customer.
Does consent shrink the audience further?
It does, and two vendors say so in their own documentation.
One publishes a regional carve-out. It “may not track onsite events for visitors to your Shopify store in the EU, EEA, UK and Switzerland, unless they have provided consent”, so the flow “will not include non-consented individuals.”
The other checks first. “Once the customer ends their session on the website, we check if the customer has emailing consent to not break the GDPR rules.”
How long should you wait before sending?
The published first-send floors run from one hour to twenty-four hours. Two platforms pick one hour; the rest differ.
One says “Add in a time delay for at least one hour.” Braze ships its template with a “1 hour delay”.
The product-side flow elsewhere goes longer, where “the minimum recommended delay for Product Abandonment is 2 hours”. Bloomreach waits three. Mailchimp offers “24 hours , 48 hours , or 72 hours after a subscriber browses your site.”
Why do the published delays disagree so much?
Because the vendors are optimising for opposite things, and both say which.
The two-hour floor exists to let the shopper carry on. It “gives contacts enough time to continue browsing and add the product to their cart before receiving a reminder.”
The one-hour floor sits inside a flow whose stated job is to reach someone while the product is still in mind.
Across the flows carrying the browse name that span is a factor of 24.
How many emails should the sequence have?
Between one and three, depending on whose guide you read, and the guides do not agree on the spacing either.
One guide says to “Send your first email within 24 hours of browsing, follow up 3-4 days later, and send a final email about a week after the initial visit.”
Another is faster and shorter. Send “within one to four hours after browsing”, follow up “24–48 hours later”, close “three to five days later”. It calls one to three emails over one to five days the ideal flow.
What does a third guide recommend?
More messages and a second channel, credited to nobody in particular.
It says that “Email marketing experts suggest building at least 3 emails and 1 SMS message into your browse abandonment flow.”
No expert is named in or beside that sentence.
How often can the same person get one?
The published hold-out runs from one day to thirty, so a regular browser is mailed far more often on one platform than another.
One builds a filter into the flow for people who “hasn’t been in this flow in the last 30 days”.
Elsewhere the shipped “default frequency of 1 day” comes with a recommendation of “10–15 days”. Mailchimp allows “one product retargeting email within a 7-day period”, and Bloomreach limits its own to “one email per 7 days.”
What happens if several abandonment flows run at once?
A shopper can qualify for four flows at once, and one vendor says that can mean 20 or more messages within 24 hours.
“A contact who browses your store, views a product, adds an item to cart, and starts checkout in a single session can qualify for all four workflows at the same time. Without the right configuration, that contact will enter every matching workflow and may receive 20 or more messages within 24 hours.”
The same page adds that “Omnisend does not have a built-in workflow priority setting.” The hierarchy is configured by hand.
Can the email show what the shopper looked at?
Sometimes. Where the trigger is a product view it can. Where the trigger is a page view, the platform never captured a product to show.
The page-view flow is blunt about the cost. “No product names, images, or prices can be dynamically pulled into the email.” And “Browse Abandonment does not store the URL the contact was viewing. There is no dynamic return URL available for this workflow.”
Its advice is to write around the gap — bestsellers, an offer, or a collection-page link.
Which product does it show if they viewed several?
One platform shows whichever product the filter you chose selects. Another always shows the last product viewed.
Klaviyo makes it a consequence of the filter chosen. With the thirty-day filter “it’ll be the first item the customer viewed”; with the other one, “it’ll be the most recent.”
The other gives one fixed answer. Its block “always shows the last product viewed by the contact”, because the trigger “captures one product view event at a time, so only one product is available in the email.”
What do the published performance numbers say?
Two vendor guides publish conversion rates for this message. They do not agree with each other, and one does not agree with itself.
The first reports that the emails “achieved a 42.16% open rate and a 0.59% conversion rate in 2026”. Its table prints the same row at 5.49% click-to-sent and 10.68% click-to-conversion.
The second reports “a 0.96% conversion rate—9.6x higher than the average email campaign”, against “the 0.10% conversion rate of the average email campaign”. Divide one by the other and 9.6 is exact. That page is dated June 2, 2022.
Do those rates trace back to anything?
Not to the report they name.
The rates are given “according to Omnisend’s 2026 report”. That report describes itself as an analysis “across 150,000 brands, based on 27 billion emails, 321 million SMS messages, and 458 million push notifications”. Its automation section reports on abandoned cart, welcome, back-in-stock and birthday messages.
The rates are labelled 2026. The report they cite measures messages sent in 2025.
Why does one guide carry two sets of numbers?
Because its prose and its table were not built from the same measurement.
The prose says “cart abandonment emails convert at 21.88%, compared to just 7.28% for browse abandonment emails. That’s three times higher, despite reaching smaller audiences”. Divide 21.88 by 7.28 and you get 3.005, so the sentence is true of its own pair.
The table on the same page prints 1.51% and 0.59% for those two message types. That ratio is 2.559. No population or period is given for either figure in the prose.
What should you check before quoting a benchmark for this email?
Check which shopper the rate counted, whether the source gives any denominator or population, and whether the source it names actually prints the figure and the period it covers.
Which shopper it counted. Product views and page views describe different people.
What population it can describe. These flows can only reach identified or subscribed people, so do not read an email benchmark as a rate over all site traffic.
Whether the named source prints the figure. And which period it covers.
Where does that leave the flow?
Worth running, and worth naming carefully before you build it.
Nothing above argues against sending it. Someone spent time in the store, whether on a product page or a collection.
What the documentation argues against is assuming the words mean one thing. Pick the trigger you want, check which of your platform’s flows carries it, and size the audience against the people it can identify.
Then measure it against your own last quarter — the same discipline an abandoned cart flow needs, and for the same reason. A number only compares to itself.