365 days, E£18.77M revenue, 5.48x ROAS — the full picture.
Not one lucky screenshot — a full year of a real DTC account, including the crashes, the diagnosis and the system that held it profitable at scale.
Anyone can show you their best month. This case study is the opposite: the maximum date range of a DTC e-commerce account I run end-to-end on Meta — prospecting, retargeting, creative testing and scaling, across 378 campaigns and a full product catalog. One year. Everything included: the peaks, the stock-outs, the checkout crisis, and the red days.
All figures pulled live from Ads Manager, maximum date range — the full dashboard is available on a screen-share call.
Why "a year at 5.48x" is harder than "a month at 12x"
The account had months at 8.29x and campaigns at 12.55x — but that's not the achievement. Consistency at volume is. Holding a blended 5.48x across E£3.42M of spend means surviving seasonality, product fatigue, platform swings and human error, over and over, without the account ever breaking for long. That takes a system, not luck:
- A workhorse engine. The core product line ran continuously all year as the account's reliable base: E£5.76M revenue from E£1.09M spend — 5.3x — 9,210 orders. Proven creatives in rotation, disciplined testing beside it (never instead of it), losers trimmed fast on multi-window reads.
- Testing that never touches the engine. New products, angles and statics got their own space at a controlled budget share. Winners graduated in; failures stayed cheap.
- Every loss converted into a rule. Over-segmentation crashes became ad-set caps. A bad rebuild became a creative-selection law. The account got harder to break every quarter.
The crash that wasn't the ads
One story from this year says more about media buying than any winning screenshot. On April 1 the account's conversion rate fell off a cliff. ROAS dropped hard, and every instinct screamed "the ads broke — rebuild everything."
I didn't touch the ads. I checked the funnel first — and found the cliff was a website checkout issue. Traffic quality and creative performance were healthy; the store had simply stopped converting. I flagged the real cause, held the engine steady, and performance restored once the site was fixed — without a pound wasted "fixing" campaigns that were never broken.
The most expensive media-buying mistake is treating a funnel problem as an ad problem. A buyer who rebuilds your account every time the store hiccups is burning your budget on the wrong fire. Diagnosis before action is the whole discipline.
What a year of this proves
The four things this account demonstrates:
- Diagnosis before action — loser vs. fatigue vs. structure vs. funnel, identified before a pound moves.
- Protect the engine, scale the proven — the workhorse never gets sacrificed for a test.
- Every down becomes a rule — the account compounds knowledge, not just revenue.
- Real, sustained, at scale — E£18.77M over a year, not one cherry-picked screenshot.
The peak this system produced — E£2.27M in a single month at 8.29x — has its own breakdown: how the account scaled to 8.29x. The crash-recovery method that protected it is in the drop-recovery playbook. Every number here is screen-shareable live on a call — no polished PDFs, the actual dashboards.