Shopify Subscription Funnel Case Study in Pakistan

The Client

The brand is an Islamabad-based direct-to-consumer specialty tea company founded by two siblings in 2021. The catalogue covers 38 SKUs across four families: single-origin black tea sourced from smallholder estates in the Shinkiari belt, green and herbal infusions, a masala chai line ground and blended in-house, and gifting tins aimed at the corporate and wedding market. A team of eleven runs operations — five in fulfilment and sourcing, two in content, the rest shared across service and finance — with performance marketing handled by a freelancer under founder supervision.

Sales run through a Shopify storefront supported by an Instagram following of roughly 61,000 that does most of the discovery work. Monthly volume sat around 2,400 orders at an average order value of PKR 2,850, with about 68% of orders on cash on delivery and the remainder split between cards and wallets. Media spend ran between PKR 1.6M and 1.9M a month — roughly PKR 1.3M on Meta and PKR 0.4M across Google Shopping and brand search — against typical monthly revenue of PKR 6.8M, spiking around Eid and the winter wedding season.

Fourteen months before the engagement, the team had launched the Monthly Chai Crate: pick any three tins, delivered monthly at PKR 2,200 — a 15% saving against one-time pricing with free delivery. It was the right product for the category. Pakistani households drink tea daily, a 200-gram tin empties in two to three weeks, and the category’s reorder behaviour is the most predictable in food and beverage. Yet the crate attached to just 6.5% of orders.

The founders’ brief to WeProms was blunt: “Our customers don’t stop drinking tea. They reorder when our ads find them again. We are paying twice for the same kitchen.” They wanted a Shopify marketing engagement that fixed the economics of acquisition — not another creative refresh.

This walkthrough is an illustrative composite built from patterns WeProms sees across Pakistani D2C food and beverage brands. The framework, sequencing, and outcome ranges are meant as a realistic reference for a growth team sanity-checking fit — not an audited result for a named company.

The Problem

The diagnosis surfaced six linked issues:

• Reorder habit existed; the offer didn’t fit the buying motion. A full 34% of one-time buyers reordered manually within 90 days — almost always through a retargeting ad or an Instagram reminder. Only 6.5% ever chose the subscription. The behaviour the crate was built to automate was already happening, expensively.
• First-order economics were negative. Blended CAC had climbed to PKR 1,520 — up roughly 18% across two quarters of Meta CPM inflation — against about PKR 1,140 of contribution margin on a typical first order. Every new customer started as a loss that only unplanned repeat behaviour recovered.
• The subscription was structurally invisible. It lived as a radio button inside the variant picker on four product pages, never in the cart drawer, never at checkout, never after purchase, and it had no dedicated landing page anywhere on the site.
• Paid bidding couldn’t see the difference between customers. Meta and Google optimised to a single Purchase event. A PKR 6,500 corporate gifting order and the first PKR 2,200 crate of a household that would renew eleven times were identical events to the algorithms.
• Prospecting spend leaked onto existing subscribers. There were no exclusion audiences, so active subscribers kept absorbing prospecting frequency the brand had already paid to earn.
• Renewal risk was unmanaged. COD refusals ran at 22% of first subscription dispatches, refusal and pause reasons went unlogged, and churn was discovered only when month-three revenue dipped.

Underneath all of it sat one question nobody in the business could answer with data: what is a tea household worth over six months, and what is it rational to pay to start one?

Phase 1 — Unit-Economics Audit and Event Plumbing (Weeks 1-2)

The first fortnight changed no ads and no pages. It settled the arithmetic every later decision would lean on.

SKU-family margin mapping. We built a landed-cost table per SKU family — sourcing, grinding and blending, packaging, courier by zone, COD fees, and gateway costs — then overlaid the 15% subscriber discount and free delivery. The picture was clear: gifting tins carried the highest gross margin but near-zero attach potential and violent seasonality, while the masala chai and everyday black tea families carried steady margins that could absorb the discount and still contribute. That single table decided where the subscription push would concentrate.

Cohort value separation. Twelve months of order history split into one-time buyers and crate subscribers. A one-time buyer produced about PKR 4,900 of six-month value across an average of 1.7 orders. A subscriber produced roughly PKR 10,600 across an average of 4.1 delivered cycles plus gifting cross-sell — a 2.2x ratio. That ratio, not any platform dashboard, became the ceiling-setter for what the brand could pay to start a subscriber.

Event taxonomy. Alongside the standard purchase event we instrumented subscription_started, subscription_renewed, subscription_paused, and subscription_cancelled with a reason code attached, aligned across Klaviyo, GA4, and Meta custom conversions, with transaction-ID deduplication so a renewal counted once. The platforms could finally bid toward the customer type the business actually wanted.

Audience hygiene. Active-subscriber and recent-purchaser exclusion lists were built and connected to every prospecting campaign, ending the quiet waste of paying to reacquire kitchens already on the books.

Question	Before Phase 1	After Phase 1
Margin by SKU family	Guessed	Mapped, including COD, gateway, and packaging costs
One-time buyer, 6-month value	Unmeasured	~PKR 4,900
Subscriber, 6-month value	Unknown	~PKR 10,600 — a 2.2x ratio
Subscription starts in ad platforms	Fused into Purchase	Dedicated subscription_started event
Renewals, pauses, cancellations	Untracked	Tracked with reason codes
Subscribers in prospecting audiences	Included	Excluded

No creative had been touched by the end of week two. But the business could now price a subscriber against a one-time buyer, and every subsequent move inherited that clarity.

Phase 2 — Subscription-First Storefront Rebuild (Weeks 3-5)

Phase 2 rebuilt the purchase path so the default choice favoured renewal wherever renewal made sense — the core of any Shopify marketing engagement in a consumables category.

Purchase-mode defaults. All twelve replenishment SKUs — the chai line, everyday black tea, the infusions — received a one-time-versus-subscribe toggle above the fold, with subscribe pre-selected and the one-time price shown beside it. Defaults do the quiet work: the buyer can switch in one tap, but the path of least resistance now leads to the crate. Gifting SKUs kept one-time as the default and received only a soft “send this monthly” suggestion, because forcing subscription framing on an Eid corporate gift would have damaged a high-margin seasonal business.

A dedicated subscription landing page. The Monthly Chai Crate got its own page for the first time: how the cadence works, skip-pause-swap controls shown explicitly, the COD confirmation rhythm explained, and the prepaid quarterly option priced at an extra 5% saving. It converted at 3.8% on the mixed paid and email traffic that reached it during the phase.

A crate builder. Instead of a fixed curation, subscribers choose their three tins and can swap monthly. Variety fatigue had been a silent churn driver; the builder turned it into a retention feature and lifted average crate margin about six points against the old fixed selection, because buyers who assemble their own crate choose more of the high-margin chai line.

Lifecycle flows in Klaviyo. Working the way lifecycle email should in a replenishment category, four flows shipped: a post-purchase attach sequence at day six (“your next crate”), a replenishment reminder at day 26–30 for one-off buyers that converted 8.4% of its recipients, a churn-risk flow triggered by failed deliveries or payment issues, and a win-back for subscribers lapsed 45 days.

A COD rhythm subscribers can live with. Because two-thirds of buyers pay on delivery, the crate was rebuilt as a confirmed cadence rather than an auto-charge: a confirmation call and WhatsApp message 48 hours before each dispatch, an easy skip option, and the prepaid quarterly crate — payable by card, JazzCash, or Easypaisa — carrying the deeper discount for buyers comfortable with mandates.

Attach surface	Mechanic	Share of new starts at week 5
Replenishment PDPs	Subscribe pre-selected	61%
Replenishment email (day 26–30)	Convert one-off buyers	13%
Cart drawer toggle	One-tap switch plus free delivery	17%
Post-purchase page	One-click conversion of the just-placed order	9%

By the end of week five, attach had moved from 6.5% to 9.4% of orders — before the paid restructure had begun, driven almost entirely by traffic the brand was already buying.

Phase 3 — Paid Restructure Around Subscription Starts (Weeks 4-8)

Phase 3 ran in parallel, restructuring the account into three campaign jobs with distinct objectives, budgets, and success definitions.

One-off prospecting stayed on Advantage+ shopping over the hero SKUs and best-seller bundles, optimised to purchase with real basket values — this is the engine for gifting-season volume and first-time discovery.

Subscription acquisition optimised to the subscription_started event once weekly start volume cleared Meta’s learning threshold, aimed at subscriber lookalikes, herbal-and-wellness interest stacks, and warm audiences from the Instagram following. Bid guidance derived from the PKR 10,600 six-month subscriber value rather than the first crate’s revenue — the specific mechanism that made paying above the old blended CAC rational for this job.

Attach retargeting chased one-off buyers 21 to 90 days after purchase with crate-forward creative, explicitly excluding active subscribers so the spend never cannibalised itself. This was the cheapest subscription-start inventory in the account, converting an audience the brand had already paid to acquire.

Creative ran on a fixed grid: brew-ritual UGC built around the doodh-patti moment, a “skip anytime” objection-handling angle that answered the biggest subscription fear directly, and price-per-cup math — roughly PKR 37 a cup against PKR 450 at a café — rendered in Urdu-first captions with English overlays. Farm and blending-room footage consistently outperformed studio product shots, a pattern we see across Pakistani food D2C.

Campaign job	Spend share before	Spend share at week 8	30-day ROAS at week 8
One-off prospecting	88%	62%	3.0x
Subscription acquisition	0%	23%	2.6x on first-cycle cash
Attach retargeting	12% generic	15%	4.4x

One reporting discipline mattered more than any of the individual moves: ROAS for the subscription campaign was reported on first-cycle cash revenue only. The 2.6x number understated the job’s true contribution — renewals land in later windows — but reporting on cash kept scaling decisions honest and stopped the account from justifying overspend with projected lifetime value it hadn’t earned yet. Blended 30-day ROAS across the account reached 2.9x by week eight.

Phase 4 — Cohort Measurement and Retention Compounding (Weeks 8-12)

The final phase built the operating rhythm that keeps these gains from decaying after the engagement window.

A cohort scoreboard. One Looker Studio dashboard, reviewed weekly: attach by SKU, blended CAC including refusal costs, ROAS by campaign job, renewal rate by cycle, and COD refusal by delivery zone. Creative retirement rules were agreed in advance — any ad below the ROAS floor after PKR 150,000 of spend paused without debate.

Churn reasons, logged and answered. The cancellation exit survey (shipped in Phase 1) finally had enough volume to read, and the top reasons each got a shipped fix:

Exit-survey reason	Share of cancellations	Fix shipped
Travelling or seasonal pause	34%	Self-serve skip and pause in the subscriber portal
Delivery day mismatch	22%	Saturday delivery-window choice at checkout
Payment friction on COD repeats	19%	Prepaid quarterly crate plus wallet mandates
Variety fatigue	17%	Swap mechanic in the crate builder
Other or unresolved	8%	Monitored, reviewed monthly

A paused subscriber is cheap to reactivate; a churned one has to be re-bought at full acquisition cost. Most pauses resumed within two cycles, and 90-day retention on new cohorts climbed from 38% to 55% — the number that makes every other number in this engagement compound.

A subscriber referral loop. A PKR 300 give-get credit ran inside the renewal confirmation, turning the most engaged subscribers into a low-cost acquisition channel. Referral-sourced starts carried no media cost and the highest cycle-two retention of any source.

Staging the season. The winter wedding and corporate-gifting calendar was mapped with crate-and-gift bundles scheduled for November and December — structured, priced, and staged, but deliberately not claimed as results inside the 90-day window.

Final Results at 90 Days

Metric	Before	Day 90	Change
Subscription attach rate	6.5% of orders	13.4%	+106% (doubled)
Subscription starts per month	156	369	+137%
Blended CAC	PKR 1,520	PKR 1,080	-29%
Blended paid ROAS (30-day)	2.4x	3.3x	+38%
90-day subscriber retention	38%	55%	+17 pts
Email and CRM revenue share	9%	16%	+7 pts
COD first-dispatch refusal rate	22%	14%	-8 pts

By week twelve, subscription cohorts carried roughly 18% of monthly revenue. That is the structural change beneath the CAC line: when close to one rupee in five arrives from households that do not need re-acquiring, blended CAC falls at flat spend — which is exactly how it reached PKR 1,080 without a meaningful budget increase.

These are illustrative outcome ranges reflecting patterns WeProms sees across Pakistani D2C food and consumables brands, not an audited result for a named company. They give a growth team a realistic shape for what a subscription-first storefront rebuild and renewal-aware paid restructure can produce.

What Made This Work

1. Unit economics were settled before tactics. The SKU-family margin table and the 2.2x subscriber-to-one-time value ratio decided where the subscription push concentrated and what a start was worth paying for. Without that arithmetic, the crate would have kept being judged on first-cycle revenue against full acquisition cost — the same lens that had already stunted it for fourteen months.
2. Defaults did more work than discounts. Pre-selecting subscribe on replenishment SKUs moved attach without spending a rupee on incentives. The buyer’s path of least resistance started pointing at renewal, and the toggle preserved choice for the one-time buyer.
3. The account stopped bidding blind and stopped paying twice. A dedicated subscription_started event gave the algorithms the customer type worth chasing, while subscriber exclusions ended prospecting frequency landing on kitchens the brand had already converted.
4. Retention work was local, not imported. The churn fixes that moved 90-day retention seventeen points — skip and pause controls, delivery-window choice, prepaid quarterly crates on wallet mandates — addressed the specific friction of Pakistani COD commerce. A generic discount ladder would have bought none of it.
5. ROAS stayed honest. Reporting the subscription campaign on first-cycle cash, with lifetime value confined to bid guidance, kept scaling decisions grounded and made the 3.3x blended number one the founders could bank, not one built on projections.

What Teams Can Apply

For Pakistani D2C brands selling consumables on Shopify:

1. Compute one-time versus subscriber value per SKU family before touching bids or discounts. The ratio sets what you can pay for a start and which products can carry a subscriber discount at all. Gifting SKUs and replenishment SKUs rarely survive the same subscription framing.
2. Make renewal the default where depletion is predictable, and never where it isn’t. A pre-selected subscribe-and-save toggle on replenishment items is the cheapest attach lever available. The same default on occasion-driven products damages conversion.
3. Give the platforms a subscription-start event and exclude your subscribers. Bidding cannot favour the customer you want if it cannot see her, and no prospecting budget should be paying to reach households already renewing.
4. Design the COD rhythm as part of the product. Confirmation calls before dispatch, skip-anywhere controls, and prepaid quarterly options on JazzCash, Easypaisa, or card mandates are subscription infrastructure in this market — not upsells.
5. Report on cash, decide on value. Use cohort lifetime value to set bid guidance, but keep the scoreboard on realised revenue. It is the only way to scale without quietly borrowing against retention you have not yet earned.

WeProms Digital has applied this subscription-first framework across Pakistani consumable brands — tea and coffee, supplements, health foods, pantry staples — and the same disciplines underpin our work across digital marketing for food brands nationwide, from Islamabad and the northern sourcing belt to the Lahore and Karachi metros.