Seveshop. The Two-Month Gold Rush on an Account That Didn't Exist.
A brand-new Romanian appliance retailer. New site, no traffic, no Google account at all. I built it from zero inside the agency of record. Sixty-two days later it had returned $51,920 at 4.37 times.
A brand-new Romanian appliance retailer. New site, no traffic, no Google account at all. I built it from zero inside the agency of record. Sixty-two days later it had returned $51,920 at 4.37 times.
The Situation Before
0x
Blended ROAS
N/A
Cost per Acquisition
N/A
Average Order Value
No Google Account at All
Nothing existed. No campaigns, no Merchant Center feed, no conversion tracking. The first impression in the account's history landed on 4 June 2026.
New Site, No Traffic
A new domain with no trading history, no audience data and no remarketing pool to build on. Google had nothing at all to optimise against on day one.
Brand Queries Unprotected
Anyone searching the brand name found competitors or nothing. The cheapest, highest-intent traffic in the category was going entirely uncaptured.
The Catalog Sat Dark
A full appliance range existed on the site with no product feed behind it. Shopping and Performance Max were both closed off completely.
No Google Account at All
Nothing existed. No campaigns, no Merchant Center feed, no conversion tracking. The first impression in the account's history landed on 4 June 2026.
Brand Queries Unprotected
Anyone searching the brand name found competitors or nothing. The cheapest, highest-intent traffic in the category was going entirely uncaptured.
New Site, No Traffic
A new domain with no trading history, no audience data and no remarketing pool to build on. Google had nothing at all to optimise against on day one.
The Catalog Sat Dark
A full appliance range existed on the site with no product feed behind it. Shopping and Performance Max were both closed off completely.
The Diagnostic
Before writing a single ad or building a single flow, we ran a full revenue leak audit across the ad account, email platform, and funnel.
Zero Coverage
An entire acquisition channel sat unused while a new store had no way to reach buyers already searching.
Zero Coverage
An entire acquisition channel sat unused while a new store had no way to reach buyers already searching.
Revenue Process
We've Built
The exact systems we put in place across ads, funnel, and email to fix the leaks.
Brand Defence First
Launched brand protection before anything else. It returned $8,888 on $201 at 44.25 times, with a $2.18 cost per order. Profitable from the first week the account existed.
PMax Product Core
Built the Merchant Center feed and launched Performance Max across the range. The broad catalog campaign returned $12,764 on $2,658 at 4.80 with a $9.59 cost per order.
Shopping Backstop
Added standard Shopping as a controllable layer under Performance Max. It returned $7,365 on $985 at 7.48, the best non-brand efficiency in the account.
The Build
We've Built
The exact systems we put in place across ads, funnel, and email to fix the leaks.
Brand Defence First
Launched brand protection before anything else. It returned $8,888 on $201 at 44.25 times, with a $2.18 cost per order. Profitable from the first week the account existed.
PMax Product Core
Built the Merchant Center feed and launched Performance Max across the range. The broad catalog campaign returned $12,764 on $2,658 at 4.80 with a $9.59 cost per order.
Shopping Backstop
Added standard Shopping as a controllable layer under Performance Max. It returned $7,365 on $985 at 7.48, the best non-brand efficiency in the account.
What Didn't Work
Three campaigns burned $1,125 between them and produced almost nothing. YouTube ran eleven days for zero primary conversions before being removed. A pet care Performance Max campaign returned 0.80, spending more than it made. Window cleaning robots managed 1.88 with a $58 cost per order against a $15 account average. We killed all three rather than waiting for them to learn.
We split vacuums into narrow Performance Max campaigns on the theory that tighter targeting beats a broad catch-all. It did the opposite. The three narrow vacuum campaigns returned 1.88, 2.81 and 4.79, while the unglamorous everything-else bucket spent far more and returned 4.80. Narrowing starved each one of the conversion volume it needed to exit learning.
Three campaigns burned $1,125 between them and produced almost nothing. YouTube ran eleven days for zero primary conversions before being removed. A pet care Performance Max campaign returned 0.80, spending more than it made. Window cleaning robots managed 1.88 with a $58 cost per order against a $15 account average. We killed all three rather than waiting for them to learn.
We split vacuums into narrow Performance Max campaigns on the theory that tighter targeting beats a broad catch-all. It did the opposite. The three narrow vacuum campaigns returned 1.88, 2.81 and 4.79, while the unglamorous everything-else bucket spent far more and returned 4.80. Narrowing starved each one of the conversion volume it needed to exit learning.
The Results
4.4x
Blended ROAS
$15
Cost per Acquisition
$66
Average Order Value
Two Months, 4.37 Times
The account returned $51,920 on $11,879 across 792 orders in its first sixty-two days, from nothing.
Brand Returned 44 Times
Brand protection returned $8,888 on $201 at a $2.18 cost per order — 1.7% of spend and 17.1% of all value.
Scaled Without Breaking
July spend ran 2.3 times June's while return rose from 3.77 to 4.60 and cost per order fell 32%.
Shopping Hit 7.48
Standard Shopping returned $7,365 on $985 at a $10.46 cost per order, capped by budget rather than demand.
Two Months, 4.37 Times
The account returned $51,920 on $11,879 across 792 orders in its first sixty-two days, from nothing.
Scaled Without Breaking
July spend ran 2.3 times June's while return rose from 3.77 to 4.60 and cost per order fell 32%.
Brand Returned 44 Times
Brand protection returned $8,888 on $201 at a $2.18 cost per order — 1.7% of spend and 17.1% of all value.
Shopping Hit 7.48
Standard Shopping returned $7,365 on $985 at a $10.46 cost per order, capped by budget rather than demand.