[UPDATE] 5.77 million lei in 60 days, from a single client.
An online store we run Meta and Google for. March 28 to May 26, 2026. 65,116 lei put into ads, 5,776,607 lei in sales, 1,716 orders. Screenshots from both accounts are below, exactly as he would show you on screen.
Sales in 60 days
Orders
ROAS Meta / Google
Invested in ads
Why an update. And why no name.
The first case study shows how you pull an account out of losses. This one shows what comes after. The same two channels, Meta and Google, left to work on an already clean structure. The numbers below are not a one-day spike we cut out. They are the total of two consecutive months.
We do not name the client. Not because we hide from the numbers, but because a store pulling nearly six million lei in two months does not want competitors staring into its account. The numbers stay exact, down to the last leu. The name stays his. If you want to verify, we open the accounts live on a call, logo on screen and all.
The window: March 28 to May 26, 2026. Two ad accounts, the exact same period. 65,116 lei put into ads. 5,776,607 lei in sales. 1,716 orders. The rest is detail.
What holds this structure up.
Google, structured by margin, not volume.
Campaigns push the products that leave profit after all costs, not the ones with the highest search volume. Bidding on clean conversions, a feed with full specs, tracking that pulls returns out of the report. That is why average order value sits above 3,000 lei, not on products sold at a loss just to inflate the conversion count.
Meta, full-funnel on the products that sell.
Google catches the person already searching for the product. Meta makes the one who was searching for nothing want it. Budget concentrated on the hero products, pixel with eventId dedup, CAPI running. On Meta you build demand from scratch, so it is normal for ROAS to be lower than Google. It still lands at 38x, which most stores do not see in a year.
Reporting on profit, not vanity ROAS.
The number we steer the account by is net profit per order, not the pretty ROAS in the dashboard. The client sits in the same spreadsheet as us, sees the same figures, decides with us. That keeps the relationship clean and the numbers honest. When you get paid from what we sell, you cannot afford to report anything but reality.
The numbers, by channel.
60 days, March 28 to May 26, 2026. No rounding in our favor. Exactly what the accounts say, visible in the screenshots below.
- Google Ads: 4,439,203 lei conversion value from 1,365 orders. 148x ROAS. Average cost per order 21.91 lei. Average order value 3,251 lei.
- Meta Ads: 1,337,404 lei in value from 351 purchases. 38x purchase ROAS. Average cost per purchase 100.28 lei.
- Combined total: 5,776,607 lei in sales from 1,716 orders, on 65,116 lei invested across the two channels.
- Google brings the volume and the value, because it catches existing demand on expensive products. Meta brings the new orders, from people who were searching for nothing. Two different roles, one objective: profit, not impressions.
The screenshots, straight from the accounts.
These two are exactly what you would see if the client took you into his accounts. Same window, March 28 to May 26, 2026. Verify anytime, on a call, with the client name on screen.
What keeps these numbers real.
- 01Average order value is high, above 3,000 lei on Google. Expensive products, healthy margin, a buyer who knows what he is paying for. That is why a ROAS that would look impossible on another store is real here, and it shows up in the till, not just the dashboard.
- 02Tracking counts real revenue, not inflated signals. Returns come out of the report, value is the real one per purchase, Meta has eventId dedup. The number you see is the number that lands in the account, not what the algorithm likes to report.
- 03Two channels that do not cannibalize each other. Google on intent, Meta on new demand. Set them to fight over the same audience and you pay twice for the same order. Here each has its role, which is why the total comes out clean.
- 04You can verify it. That is all that matters in a case study. On a call we open the accounts live. If the numbers were not real, we could not make the offer we make: you pay from what we sell, not upfront, on promises.