Scaled DTC store to $2.1M ARR on profit, not vanity ROAS

A multi-country DTC apparel + wellness portfolio scaled from $180K to $640K monthly revenue at a 3.4x blended POAS by rebuilding measurement, tiering PMax by margin, and layering Klaviyo + WhatsApp on top of paid. Profit went positive in week 7, not when the founders thought it would.

The Brutal Problem

Two co-founders, three Shopify stores, one shared bank account that had been red for eleven consecutive weeks. The portfolio had crossed $2M ARR across MENA, UK and US audiences and the founders had assumed the next jump to $5M was a budget problem — pour another $40K/month into Meta and the math would follow. Instead, every spend increase made cash worse. They had stopped paying themselves in March. One co-founder's wife was six months pregnant and they'd had a very quiet conversation about whether the family could absorb a year without his salary. The other co-founder's parents had asked, gently, whether he could move back into their spare room for a quarter. They had laid off their only in-house creative in April and were running ads on assets eight months old. The Meta dashboard claimed a 2.6x ROAS portfolio-wide; the bank account said something closer to break-even minus shipping. Customer reviews were strong (4.6 average), repeat-intent surveys good, NPS healthy. Product wasn't the problem. Acquisition wasn't even the problem. The problem was that nobody could tell which dollar of spend had made money and which had set fire to itself, and the founders were 90 days from selling one of the brands at a fraction of valuation just to keep the others alive.

What Made It Worse

Five cascading failures. (1) Browser-only pixel across all three stores — Event Match Quality stuck in the 4s. (2) PMax was scaling the three highest-revenue SKUs which happened to be the three lowest-margin. (3) Creative refresh had effectively stopped after the in-house hire was laid off. (4) Klaviyo flows existed but were portfolio-generic, not store-segmented. (5) The shared Meta Business Manager was leaking learning between accounts — retargeting audiences and prospecting audiences shared learning pools across all three brands.

The Diagnosis

Three-store audit produced one identical answer. EMQ across all stores averaged 4.4. True purchase capture was 61% of actual orders — Meta was blind to almost 40% of revenue. Portfolio reported ROAS 2.6x, POAS-adjusted (margin, shipping, returns) was 0.88x. PMax allocation: 71% of spend on bottom-quartile-margin SKUs. Klaviyo revenue share was 6.3% portfolio-wide vs benchmark 25-35%. Creative fatigue: top-spend ads at frequency 5.1+, CTR collapsed 47% in the prior 60 days. The diagnosis: the portfolio was burning ~$71K/month in pure inefficiency — wrong signal, wrong SKUs, wrong assets, wrong owned-channel coverage. Fixing all four in sequence would flip cash positive without an additional dollar of spend.

The Solution Stack

12 weeks, 11 steps:

1. Week 1 — Per-store Stape CAPI gateways + dedicated sGTM containers. Isolated learning per brand.
2. Week 1 — Deduplication + EMQ verification — all three stores climbed past 8.0 inside ten days.
3. Week 2 — Margin + return-rate feeds piped into Meta and Google per store.
4. Week 3 — PMax margin tiers — premium / mid / clearance fenced separately. Low-POAS SKUs blocked from scaling.
5. Week 4 — Creative Velocity restart — 18 variants/month/store via a freelance creative pod, 72-hour kill window.
6. Week 5 — Klaviyo per-store rebuild — segmentation, welcome, AC, post-purchase, win-back, VIP per brand.
7. Week 6 — WhatsApp recovery on MENA store, SMS on UK/US.
8. Week 7 — Audience hygiene — isolated retargeting from prospecting per account.
9. Week 9 — Portfolio profit dashboard reconciled daily to bank deposits.
10. Week 10 — Cross-store learnings — top creative concepts adapted across brands.
11. Week 12 — Spend reallocation — portfolio budget rebalanced to highest-POAS store with capacity.

The Inflection Point

Week 7. The POAS feeds had been live for three weeks across all stores and the algorithm visibly reweighted overnight on the Sunday. The MENA store, which had been the perceived "weakest" by reported ROAS, jumped to portfolio-leading POAS as its margin-tiered PMax started compounding. The UK store's Klaviyo rebuild went live the same week and added 11% of revenue in seven days. Monday morning the founders opened the new dashboard and saw the portfolio sitting at $19K positive contribution margin for the week — the first positive week in a quarter. One of them screenshotted it. The other called his wife.

Final Numbers

Metric	Before	After	Change
Monthly revenue	$180K	$640K	+255%
Portfolio POAS	0.88x	3.4x	+286%
EMQ (avg)	4.4	8.2	+86%
Klaviyo + SMS share	6.3%	31%	+4.9x
Blended CAC	$34	$19	-44%
Contribution margin	-9%	+22%	+31 pts
Monthly net cash	-$24K	+$118K	flipped

What We Learned / Replicable Playbook

Multi-store DTC portfolios fail the same way single stores do, only louder. The replicable sequence applies regardless of count: (1) isolate measurement per store — shared BM creates shared signal contamination, (2) verify EMQ above 7 per property before judging any campaign, (3) feed margin and return-rate per SKU per store, (4) tier PMax by margin individually per brand, (5) build Klaviyo per store, not portfolio-generic, (6) only then rebalance spend across brands. Portfolio operators who treat all stores as one media account starve their winners and overfund their losers — the math compounds in both directions. Isolation first. Compounding second.

Daily-Life Outcome

Both co-founders restored their salaries in month four. The pregnant co-founder's family stayed put; he was at the hospital for the birth without a laptop. The other moved into a one-bedroom of his own in month five. They rehired their creative lead in month six, and added a junior performance marketer in month eight. The "should we sell a brand" conversation stopped coming up. They're planning a Saudi market entry for Q2 2027.

FAQ

Does this work for a single Shopify store?

The same sequence applies, faster — typically 8-10 weeks vs 12.

How much creative volume is actually needed?

10-18 variants per store per month is the sweet spot for sub-$1M/month spend brands. More than that produces diminishing returns without a dedicated CRO loop.

What if our margins are too thin for POAS bidding to work?

Then POAS bidding is the only thing that will keep you alive. Thin margins amplify the cost of bidding to revenue.

Driving services: DTC Ecommerce Growth · Server-Side Tracking · SEM & Paid Advertising. Further reading: DTC Scaling on Profit · POAS vs ROAS.

Get a free 48-hour audit → We'll quantify which of your DTC stores is leaking margin — and the exact sequence to flip it positive.