Browse Abandonment

Look at this journey when product analytics show many users opening the same product several times and leaving without adding it to the cart. You need product views counted per user and product, add-to-cart and purchase tracked on the same identity, push permission recorded, and a response window set to fit your business. If the repeated views come from a product page problem such as missing size details, an unclear shipping cost or images that fail to load, fix the page first; a push only sends the user back to the same uncertainty.

Incremental viewed-product purchase rate vs control

The single push is worth keeping only if the reminded group buys the product it viewed more often than the control group; a tapped push or a reopened product page is not the outcome.

How it is counted: Take all users who met the trigger condition and were assigned to either group, and count the share who buy the viewed product within the same follow-up window from assignment; the gap between the groups is the journey's effect. Whether the push arrived or was tapped does not change the denominator.

Also track

• Add-to-cart rateAssigned users who add the product to the cart after the push, as a share of everyone assigned; the intermediate step on the way to purchase, and where the handoff to Cart Reminder takes place. Measure it in the control group as well.
• Converted before the pushThose who add to cart or buy during the 6 to 24 hour wait and leave without a message; they convert on their own, and a push would have been noise for them.
• Push opt-outsUsers who turn off notifications after the push; interest inferred from views is a weaker signal than a saved item, so tolerance for the message may be lower.
• No-action exitsEntries whose response window closes with no action; in this flow a no-action ending is the common outcome and is not a failure by itself.

• Trigger and entryEntry requires the same product viewed at least twice with no add-to-cart or purchase. Split the users who meet that condition at random: the journey group gets the reminder, the control group sees no push for this product.
• Who must not enterAnyone who already added the product to the cart or bought it stays out. A user who saved the product belongs to Saved Product Reminder, one who asked for a stock alert to Back-in-Stock Alert, which suppresses this flow for the people it holds. Scope is one viewed product, with no second flow for it.
• Timing and channelsAfter a wait of 6 to 24 hours, add-to-cart and purchase are checked again; if neither happened, a single push goes out. The push recalls only the viewed product and creates no urgency without real stock, price or promotion data. The journey then waits through the response window you set for your business.
• Exit and handoffAn add-to-cart or purchase at the first check, or a purchase after the push, ends the journey. If the user adds the product to the cart after the push, the user moves to Cart Reminder and this sequence closes; when the response window passes with no action, the journey ends without a second reminder.
• Frequency and consentOnly one push per viewed product, sent only to users with notification permission. Since push is the only channel, a user without it never enters; keep browse and cart messages for the same product from stacking up.

1. 1A push that leaves the journey group's purchase rate level with the control group's only reached users who were going to convert anyway; tap counts or returns to the product page do not offset that. Base the decision on how much the two purchase rates differ, not on views.
2. 2What you measured belongs to this single push, the 6 to 24 hour wait and the response window you chose. In journey reports, a purchase by a user who added to cart after the push and moved to Cart Reminder is credited to that journey; in the control comparison, count purchases for both groups by one and the same rule.
3. 3Do not expect a large gap from one push; seeing a small gap with confidence takes many entries. Enter your current purchase rate and the smallest uplift you want to detect into the sample size calculator to get the users needed per group, then divide by your daily eligible entries to see how many days the test must run.
4. 4If the push used urgency wording such as limited stock or a time-limited discount without data behind it, do not count the resulting sales as the journey's gain; tie the copy to real stock, price or promotion data first.