Browse Abandonment Emails: Bringing Window Shoppers Back
A weaker signal than a cart, which changes the tone entirely — and the threshold that stops it feeling like surveillance.
On this page
The short version
- A view is a much weaker signal than a cart, and treating it like one produces messages that read as surveillance.
- Set an intent threshold — repeat views, dwell time, several items in one category — rather than firing on any page view.
- Reference the category rather than the exact item where you can. It is nearly as useful and considerably less unsettling.
Browse abandonment applies cart recovery's logic to a weaker signal: somebody looked at something and left. It can work, and it fails in a way cart recovery does not — by making the recipient uncomfortable.
The difference is what the signal means. Adding to a cart is a decision; looking at a page is often idle. An email treating the second as though it were the first is claiming a level of interest the person did not express.
The intent threshold
Firing on any single page view produces a large volume of low-quality sends and a reasonable number of people wondering how closely they are being watched.
Something more is needed before the signal is worth acting on: the same item viewed more than once, several items in one category in a session, a long dwell on a product page, or a return visit to the same item on a different day.
Each of those is a genuine expression of interest rather than a passing look, and each raises both the conversion rate and the comfort of the recipient — because a message about something they clearly considered reads differently from one about something they glanced at.
Signal strength by behaviour
| Behaviour | Signal | Send? |
|---|---|---|
| Same item viewed on two separate days | Strong | Yes |
| Several items in one category, one session | Strong | Yes, about the category |
| Long dwell on one product page | Moderate | Yes, gently |
| Used a filter or comparison tool | Moderate | Yes |
| One product page view | Weak | No |
| Category page only | Very weak | No |
| Arrived and left within seconds | None | No |
Tone, which differs from cart recovery
A cart email can reasonably say "you left something behind", because they did. A browse email cannot, because they did not leave anything anywhere.
The register that works is recommendation rather than reminder: here is something in this area, here are two others like it, here is what people usually ask about them. It is closer to merchandising than to recovery.
That also makes it more useful when the specific item was wrong. A reader who looked and decided against is still a reader interested in the category, and a recommendation serves them where a reminder about the rejected item does not.
One message, usually
Cart recovery justifies three because the intent was explicit. Browse abandonment justifies one, sent within a day, because the intent was inferred.
A second message about something someone merely looked at crosses from helpful into insistent, and the incremental conversion does not cover the unsubscribes.
Where the category signal is strong — several items, repeated sessions — a second message about the category rather than the item is defensible. The item itself gets one mention.
Who to exclude
Anyone who bought the item or anything in the category since. Anyone currently in a cart recovery flow, because the cart is the stronger signal and two flows about the same session is the collision that reads worst.
Anyone in their first days as a subscriber, who has not yet had a normal email from you — a browse email as the second thing you ever send is a strange introduction.
And anyone who has told you they want less. Browse abandonment is the most easily skipped flow when honouring a frequency preference, because its value per message is the lowest of the commercial flows.
Browse abandonment check
- An intent threshold beyond a single page view
- One message, sent within a day
- Category framing rather than a specific observation
- No timestamps or session details in the copy
- Suppressed if a cart flow is running
- Suppressed for new subscribers
- Suppressed for anyone who bought in that category
- Skipped first when honouring a reduced-frequency preference
Frequently asked questions
Does browse abandonment work as well as cart recovery?
No, and it is not close — the signal is weaker and the audience larger. It is worth having after the higher-return flows exist, not before.
How long after browsing should it send?
Within a day, and not within the hour. An email arriving minutes after someone closed a tab is the version that reads as watching rather than helping.
Should it work for content sites without products?
The equivalent — recommending more on a subject someone read repeatedly — works well and reads better, because content recommendations are an expected part of a publication.