a Danish travel brand launched with no online revenue and built the whole business on paid. Twice the ads came off, once through a policy takedown. Store revenue fell from about 88,000 DKK on a peak day to about 2,000 DKK, sometimes negative once returns were counted. Facebook was crediting itself with about 70% of revenue. The days with nothing running put the real figure at 95 to 99%. It was an accident rather than a designed holdout, and it is still the finding Simon uses on other accounts to argue for switching things off before believing an attribution report.
Key results.
These are store revenue figures either side of an unplanned ad stop in roughly October to December 2024, described by Simon on a recorded call. They are a peak day and an off day, not a run rate, and not a controlled incrementality test.
Challenge
SafeCarry was new. In Simon’s words on a later sales call, “Safe Carry was completely new.” There was no online revenue to build on and no purchase history for the pixel to learn from, so every sale in the first months had to be bought outright.
The market had a hard edge on it. One country, one language, one core product line, and no international expansion at any point in the engagement. A Danish brand selling only in Denmark runs out of addressable audience at a size most brands never have to think about.
The plumbing was not ready for spend either. Before any Google budget went live, conversion tracking had to be confirmed, which is what held the Performance Max build in July 2024. The product feed had been handed over as a static file rather than a live link, which is enough on its own to stop Shopping and Performance Max working properly.
And nobody knew what the advertising was actually worth. Facebook reported a share of revenue in the region of 70%. That number was the only answer anyone had, and there was no reason to trust it in either direction.
Solution
Step #1. Fix what the spend depends on, before the spend.
Google was gated on tracking verification rather than launched on a schedule. The feed was moved off a static file onto an XML link in September 2024 so Merchant Center had something live to read.
Step #2. Build the creative engine in-house rather than buying it.
Video scripts were written in-house in Danish, and video ads went live on Meta off client-supplied and UGC footage. The best-performing asset, a Cozy Pillow UGC clip, was re-edited to push it further. The instructions Simon wrote to help the founder film usable footage were good enough that the team adopted them internally as the standard for every client founder we ask to shoot.
Step #3. Treat the ad-off days as evidence instead of as lost revenue.
When the ads stopped, the store went from about 88,000 DKK on a peak day to about 2,000 DKK, sometimes negative once returns landed. Against a reported 70%, the observed contribution of paid was 95 to 99%. Growth Mate has argued from that gap on other accounts ever since, with the caveat stated every time: it was a policy takedown, not a holdout anyone designed.
Step #4. When the country ran out, change the creative rather than the budget.
By March 2025 the account had reached 1.5 million people, which Simon put at about 30% of Denmark’s population. The account was decaying because there was nobody left who had not seen the ads. The instruction that went out was new angles into problem-aware audiences, not more spend behind the same creative.
Alongside the ad work, Growth Mate set up the order-confirmation and daily-orders email so the founder stopped having to be sent his own numbers, fixed slider animation, logo spacing, form design and a pop-up on the Shopify theme, and scoped a branded-versus-current site A/B test in April 2025.
Results
Book a discovery call about your business goals
- ✓Non-binding introductory meeting
- ✓Explore your growth potentials
- ✓Talk to our experts