Browse and Cart Recovery for an Electronics Retailer

The Client

A Multan-based consumer electronics and home appliances retailer with four showrooms across South Punjab and a growing online store. The family-run business had been selling televisions, refrigerators, air conditioners, and mobile phones for over a decade, and had launched its ecommerce site two years earlier to capture demand beyond walk-in customers across Pakistan’s broader electronics retail market.

Average order value online was high — roughly PKR 54,000 — because customers mostly bought considered, big-ticket items rather than impulse purchases. That considered-purchase dynamic sat at the heart of the business’s online problem: shoppers added a refrigerator or an LED TV to their cart, then left to compare prices on Daraz, check with a family member, or wait for the next paycheck.

The online store ran alongside four physical showrooms, and the two channels shared inventory but not data. Showroom staff closed sales face-to-face, but online shoppers received no such intervention — once they left the site, they were gone. Cash-on-delivery dominated online orders (roughly 78% of completed purchases), which meant the final commitment often happened days after the cart was built, leaving a long window in which a buyer could drift to a competitor or defer the purchase entirely. Driving more traffic would have widened the same leak; the cheaper, higher-leverage move was to recover the demand already arriving and then leaving.

The team had email marketing running through a basic setup — a monthly newsletter and a single generic abandoned-cart email sent 24 hours after abandonment — but it contributed almost nothing to revenue, and no one on the small team had the time or tooling to build anything more sophisticated.

The Problem

Online revenue had plateaued around PKR 22 million per month, and the gap between traffic and orders was widening. The diagnostic showed the leakage was concentrated at two points.

• Cart abandonment at 78%. Roughly four in five carts were abandoned, in line with electronics norms but leaving a large unrecovered pool of high-intent buyers.
• Browse abandonment unaddressed entirely. Visitors who viewed a product three or more times without adding to cart received nothing — no nudge, no price signal — even though this behavior strongly predicted a near-term purchase in a comparison-heavy category.
• The single cart email earned PKR 180,000 a month. Under 1% of recovered revenue, against an industry expectation of 10-15% from a complete recovery program.
• No product or price context in recovery messages. The generic cart email linked to the cart but carried no stock urgency, no price-drop signal, and no pickup-in-store option, which mattered because many Multan buyers prefer collecting big-ticket items from a known showroom.
• Email captured at checkout only. Because the email field sat inside checkout, the store could only message shoppers who had already started entering their details, missing the large pool who browsed or added to cart anonymously.
• Email deliverability unmonitored. The basic setup had no sender authentication or inbox-placement tracking, so a meaningful share of the few recovery emails sent never reached the inbox.

The comparison-shopping behavior was the underlying driver. A buyer researching a PKR 145,000 inverter AC would open the store’s product page, then three competitor pages, then return two days later to re-check, then wait for a salary cycle. These were not lost buyers — they were buyers mid-decision — and the store’s existing setup treated every one of them as if the decision had already been made (a single generic cart email) or as if no decision was happening at all (silent on browse abandonment). The mismatch between buyer stage and message stage was where the revenue was leaking. The store was leaving high-intent, high-AOV demand on the table at exactly the two moments — browse and cart — where a well-timed message most often changes the outcome, which is the core of a complete browse and cart recovery program.

Phase 1 — Capture, Tracking, and Deliverability Foundations (Weeks 1-2)

Recovery only works if you can identify abandonment, reach the shopper, and measure attribution. The first two weeks were foundational rather than creative.

• Email capture moved earlier. We added an email-capture step at add-to-cart and on the cart page itself (offering a small price-drop alert opt-in), so the store could message browsers and cart-adders who never reached checkout. Capture rose from roughly 40% of purchasers to 71%.
• Behavioral tracking layered in. We instrumented three events: product_viewed (with repeat-view detection), cart_added, and checkout_started. These events became the triggers for the recovery flows.
• Email deliverability hardened. We authenticated the sending domain (SPF, DKIM, DMARC), warmed a dedicated sending IP, and added inbox-placement monitoring. Open rates on the existing newsletter rose from 14% to 27% purely from better deliverability, before any new flow launched.
• Attribution set to a windowed last-touch model. Recovery-attributed revenue was tracked against a 7-day window for browse flows and a 14-day window for cart flows, with a cap so recovery could not claim revenue from shoppers who would have converted anyway.

The capture redesign mattered more than it looked. The add-to-cart email prompt was framed as a utility — “Tell me if this drops in price” — rather than a newsletter subscription, which lifted opt-in threefold versus the earlier checkout-only field. On deliverability, the DMARC policy was rolled out in monitoring mode first (p=none) for ten days to catch authentication failures without bouncing legitimate mail, then escalated to quarantine once the domain’s sending reputation stabilized. This sequencing is what let the newsletter open rate climb to 27% without a deliverability incident. This phase produced no revenue on its own, but it made every later number trustworthy.

Phase 2 — Browse and Cart Recovery Flow Design (Weeks 3-5)

With triggers and attribution in place, we built the recovery program as two distinct lifecycle email flows, each tuned to how electronics buyers actually decide.

Browse abandonment flow (three emails plus one onsite). Triggered when a shopper viewed the same product three times in seven days without adding to cart.

Step	Timing	Content	Goal
1	Day 0, +2 hrs	”Still comparing?” — the product, a one-line spec summary, and a price-drop opt-in	Capture intent and opt-in
2	Day 2	Social proof — ratings, a short review excerpt, and showroom pickup availability	Reduce comparison uncertainty
3	Day 4	Stock urgency — live stock count for that SKU at the nearest showroom	Convert the deliberate buyer
Onsite	Real time	A soft banner on return visits showing the last-viewed product with stock and price status	Re-engage without email

Cart abandonment flow (three emails plus WhatsApp). Triggered at cart abandonment, segmented by cart value.

Step	Timing	Content
1	+1 hr	Cart contents with a direct checkout link and the showroom pickup option
2	+24 hrs	For carts over PKR 80,000 — an installment-plan summary and warranty highlights; below that — a comparison guide versus the runner-up model
3	+3 days	A final nudge with genuine stock urgency (“2 left at the Multan Boulevard Road showroom”) and a one-tap WhatsApp reply to reserve

We deliberately avoided percentage discounts as the default lever, because aggressive discounting erodes margin on already-thin electronics markup and trains buyers to wait. Instead, the flows leaned on stock urgency, installment options, warranty reassurance, and pickup convenience — the levers that actually move a considered electronics purchase in Pakistan.

The relative size of the two audiences surprised the team. Because email capture now happened at add-to-cart, the browse-abandonment audience (repeat viewers who never added to cart) was more than four times larger than the cart-abandonment audience. The store had been ignoring the larger pool entirely. Browse flows, once live, contributed roughly 35% of recovery-attributed revenue in the first month, and the gap narrowed only as cart flows matured — a useful reminder that “abandoned cart” and “abandoned browse” are not the same problem and do not reward the same solution.

Phase 3 — High-Value Cart Handling and Channel Mix (Weeks 4-8)

Because the store’s AOV was high and big-ticket carts behaved differently from accessory carts, we split cart recovery by value and added a channel beyond email.

• Big-ticket branch (PKR 80,000+). These carts — refrigerators, ACs, large TVs — got the installment-plan and warranty-led second email, plus a WhatsApp message from the showroom team offering to hold the unit for 24 hours. WhatsApp carried a human name and a real showroom line, not a bot.
• Standard branch (under PKR 80,000). Smaller electronics and accessories got the comparison-guide second email, leaning on review and spec comparison rather than financing.
• Price-drop alerts wired to browse and cart. When a viewed or carted SKU dropped in price, a one-off price-drop email fired within the hour, tagged so it would not clash with the scheduled flow. Electronics buyers are highly price-sensitive, and a real price drop — not an invented one — converted reliably.

The installment-plan branch leaned on a partner financing option the store already offered in showrooms but had never surfaced online. Showing “PKR 12,100/month for 12 months” beside the full price on the recovery email converted buyers whose blocker was lump-sum affordability, not product fit. That single line item recovered a category of high-AOV carts that flat discounting could never have reached profitably, because it addressed the actual constraint (cash flow) without eroding the unit margin the store protected by holding price.

The WhatsApp addition mattered more than expected. For high-AOV items, many buyers had a final question (delivery to their town, installation, exchange of an old unit) that an email could not answer fast enough. A WhatsApp line staffed during showroom hours let the recovery flow hand off to a human at the decision moment, which is exactly where big-ticket carts typically stall.

Phase 4 — Tuning, Suppression, and Compounding (Weeks 8-12)

With the flows live and attributed, we spent the final phase tightening the program so it compounded without annoying shoppers.

• Suppression rules. Anyone who purchased, or who had been emailed three times in the cycle, was suppressed to prevent fatigue. Repeat shoppers entered a lighter post-purchase path instead of recovery.
• Stock-urgency accuracy. The stock-count language was tied to live inventory so urgency was always truthful — invented scarcity destroys trust in a category where buyers comparison-shop across five sites.
• Send-time optimization. Browse and cart emails shifted to late evening (8-10 pm), when Pakistani shoppers comparison-shop after work. Open and click rates rose roughly a third versus the original daytime sends.
• A/B testing subject lines and the WhatsApp handoff. The single biggest cart-recovery lift came from moving the pickup-in-store option above the checkout button in email one, because a meaningful share of Multan buyers prefer collecting a big-ticket item themselves.

By week 12, recovery-attributed revenue had reached roughly PKR 3.1 million per month — about 14% of the store’s monthly online revenue — and total email revenue (recovery plus the existing newsletter and welcome flow) had risen from under 7% to roughly 21% of online revenue.

By attribution, the WhatsApp handoff and the big-ticket installment branch together drove the majority of recovery revenue, because they converted the carts with the highest order value — the ones where a recovered sale mattered most to the monthly total. Browse flows contributed the largest count of recovered orders but at lower average value, while the deliverability work amplified every flow by lifting the share of messages that actually reached the inbox. The lesson the team took from the attribution split was to weight effort by revenue contribution per recovered order, not by raw recovered-order count.

Final Results

At 90 days, the recovery program had become one of the store’s largest revenue channels.

Metric	Before	After	Change
Recovery-attributed revenue / month	PKR 180,000	PKR 3.1M	+14% of monthly revenue
Cart abandonment recovery rate	4.2%	12.8%	+8.6pp
Email revenue share of online total	~7%	~21%	+14pp
Email open rate	14%	31%	+17pp
Average recovered order value	—	PKR 54,000	—

The store expanded the program in the following quarter to a win-back flow for lapsed buyers and a price-drop subscription across the catalog, projecting recovery revenue to compound toward 18-20% of online revenue without further ad spend.

What Made This Work

1. Capture moved before the message. The single biggest unlock was capturing email at add-to-cart rather than at checkout. It roughly doubled the addressable audience for recovery, and every later flow depended on it.

2. Browse and cart were treated as different decisions. A browser is comparing; a cart-adder has chosen and is stalling. The two flows used different levers — comparison proof versus stock urgency and pickup — because the buyer psychology is different.

3. Urgency was real, not manufactured. In a category where buyers check five sites and remember prices, fake scarcity or phantom discounts get caught and destroy trust. Stock counts and price-drop alerts were tied to live data, which made urgency credible and effective.

4. WhatsApp closed the high-AOV gap. Email alone could not answer the final question on a PKR 120,000 refrigerator. A staffed WhatsApp line handed the recovery flow off to a human at the decision moment, which is where big-ticket carts die.

5. Deliverability was treated as infrastructure. The new flows would have underperformed by a third without the sender-authentication and inbox-placement work done in week one. Recovery only works if the message lands in the inbox.

What Teams Can Apply

These takeaways apply to any Pakistani high-AOV ecommerce store where buyers research, compare, and defer before committing — electronics, appliances, furniture, and jewelry all share the considered-purchase shape that makes recovery the highest-leverage channel you can stand up before buying any new traffic. The compounding effect is the real prize: recovery revenue grows month over month as the flow library matures, without a proportional rise in ad spend.

1. Capture email before checkout. Move your email field to add-to-cart or the cart page with a price-drop opt-in. You cannot recover shoppers you cannot identify, and most electronics abandonment happens before checkout.

2. Separate browse and cart recovery. Build a browse flow for comparison-stage visitors and a cart flow for chosen-and-stalling buyers, each with its own timing and levers. A single generic cart email leaves the browse pool entirely untouched.

3. Use real urgency, never invented scarcity. Tie stock counts and price-drop alerts to live data. In a price-transparent category, honest urgency converts and dishonest urgency burns the brand.

4. Add a human channel for high-AOV carts. For carts above your category’s considered-purchase threshold, a WhatsApp or call handoff from the recovery flow answers the final blocking question that email cannot. This is where most big-ticket recovery actually happens.

5. Fix deliverability before you build flows. Authenticate your sending domain, warm the IP, and monitor inbox placement first. The most sophisticated recovery sequence earns nothing if it lands in spam.