The abandoned browse playbook
Ten times more sessions end in browsing than in carts. A browse flow done badly is surveillance email; done well, it reads as a helpful nudge about something the shopper genuinely considered.
Browse abandonment is the largest recoverable audience in the store and the easiest flow to get wrong — the difference between "helpful reminder" and "we are watching you" is intent thresholds, message framing, and restraint. This playbook sets the signals that distinguish real consideration from idle scrolling, writes messages that help rather than surveil, and caps the machine so it never becomes the reason someone unsubscribes.
Stage 1: Define real intent, not a pageview
Triggering on a single product view mails idle scrollers and reads as creepy. Real consideration looks different: repeated views of the same product, meaningful dwell, image gallery engagement, size-guide or shipping-info visits, or return visits to the same category within days. Set the threshold from combinations of these — the flow should fire on the sessions that resemble your actual pre-purchase behaviour, discoverable from your own funnel data.
Trigger on signal combinations: repeat views, dwell, detail engagement — never a lone pageview
Threshold calibrated against your own pre-purchase session patterns
Suppression for recent purchasers of the category and active cart holders
Identity handled honestly: only mail identified, consented subscribers
Stage 2: Write the nudge, not the surveillance report
The framing rule: reference the category, feature the product. "Still thinking about trail runners?" with the viewed product and two alternatives reads as service; "You looked at this exact item at 9:47" reads as a warning label. Content that helps the consideration continue — the size guide, the review highlights, the shipping answer — outperforms bare product-plus-discount, and holds the incentive back entirely: browse intent is too early for coupons, and training browsers to expect them is expensive.
Stage 3: Cap it, then judge it honestly
Browse triggers fire constantly, so the caps do the work: one browse email per cycle, hard frequency ceilings across all flows combined, cart and post-purchase flows always taking precedence, and automatic suppression once the item is bought. Judge the flow on conversion and unsubscribe rate together — a browse flow with strong revenue and rising unsubscribes is borrowing against the list, and the caps need tightening.
How it works
Calibrate the trigger
Intent thresholds from signal combinations, matched to real pre-purchase patterns.
Write category-first helpers
The considered product plus alternatives and genuinely useful content — no incentives.
Cap and monitor both metrics
Frequency ceilings across flows, precedence rules, and conversion judged alongside unsubscribes.
What you get
A flow that fires on consideration, not on curiosity
Messages that read as help and convert like it
Frequency discipline that protects the whole programme
The largest recoverable audience actually recovered
Handled by these agents
Works with
Frequently asked questions
How is this different from cart abandonment?
Intent depth. A cart is a decision nearly made; a browse is a consideration forming. The browse flow therefore helps the decision along rather than pushing completion — different timing, different content, and no incentives.
Will customers find browse emails creepy?
Category-framed, consented, and frequency-capped: no — it reads like a shop assistant remembering you. Product-timestamped and relentless: yes. The framing and caps in this playbook are exactly that line.
What conversion should a browse flow produce?
Below cart recovery per send — intent is earlier — but the audience is many times larger, so aggregate contribution often rivals it. Judge on segment revenue and list health together, never per-send rates alone.
Which agents run this?
Analyst calibrates thresholds from session data; Retention writes and runs the flow with the caps enforced account-wide. Suppression coordination across flows lives in the shared layer, where it cannot be forgotten.
Keep reading
Reduce cart abandonment
A diagnostic-first programme: find why carts die, fix the causes, then automate recovery.
Email flow automation
Complete lifecycle coverage, built once and then actually maintained.
Customer segmentation
Segments built from purchase behaviour, not from generic engagement windows.
Measure marketing attribution
An honest attribution setup for stores: clean data, triangulated methods, stated uncertainty.
Put the team to work on your store
Connect your storefront, analytics, and email stack, set a goal, and let the agents run the work end to end. Start on the free plan with your own model key.