From €8K to €45K a day without losing the margin

01The starting pointWhere the brand was stuck

The product sold. The ads worked. The problem was everything around them.

When we started, the brand was spending €8K a day at 2.3× ROAS. Every attempt to push past that level ended the same way. ROAS slipped, the same few ads wore out, and more than once an account got restricted along the way.

In the 4 months before we started, they had 2 accounts restricted and €103K in spend frozen during appeals. After a few of those weeks, the team stopped planning bigger budgets.

02EconomicsThe numbers we agreed before touching a campaign

Scaling a supplement brand starts with what a new customer is allowed to cost. If the team doesn't agree on that, every budget review turns into an argument about ROAS screenshots.

We built a simple first-order model with the founder:

Two rules came out of it. First, we judged everything on first-order economics, never on hoped-for repeat purchases the brand hadn't proven yet. Second, the lead offer became the three-pack: 3 × 30 servings; €100 average net order value, because it was the only SKU with enough margin to carry acquisition at scale.

03The rebuildWhat we changed

No single fix moved this account. The results came from creative, structure, testing and the account setup changing together, each one making the others work better.

Swipe the table sideways to see every column →

Before raising budgets, we moved the brand onto our agency ad accounts and set up backups with the main campaigns mirrored. New ads and landing pages went through a compliance check before launch, to identify potential issues before launch.

Funding ran through top-ups. The brand kept about 7 days of spend on the balance and topped up ahead of each budget step.

The next four sections go through each change in detail.

04CreativeThree buying reasons, not thirty variations

The old creative library was 12 ads that were mostly the same video with different text overlays. They relied on the same buying reason, and when that message fatigued, the account had nothing left.

We rebuilt it around three distinct reasons someone buys a supplement. Each one answers a different question in the buyer's head:

Concept · Routine

"My daily supp stack."

A native notes-app list of a real daily routine: collagen in the morning coffee, organs with breakfast, protein after training. It answers will I actually use this?

Format: Static 9:16; notes-app screenshot

Concept · Transparency

"Nature's original multivitamin."

A product static that calls out what is in the capsule, the daily dose and where the cattle come from. It answers can I trust what's in it?

Format: Static 1:1; ingredient callouts

Concept · Pack choice

"Protein, collagen and organs. The full stack."

A natural flat-lay of the three core products together, so the buyer sees the complete stack at once. It answers which one should I get?

Format: Static 4:5; product flat-lay

What didn't work

Founder-story video, €480 spend and 5 new customers (€96 CPA). It explained the company but left pack choice and price unclear; stopped after the matured review.

The production rhythm

Volume only helps if the team can turn it into decisions. We settled on a pace the brand could sustain every week without burning out creators or the buyer:

• 3 new concepts per week, each with 2 alternate openings
• One static treatment for every video concept
• Every brief names the buying reason, the proof on the label, and the exact landing-page section the ad sends people to
• No health outcomes, before/afters or invented testimonials. Creators only show their real use.

Each ad landed on the matching part of the product page. The routine static went straight to the usage section, the transparency static to the ingredient panel, and the pack static to the pack selector. The page finished the job the ad started.

05StructureEvery campaign got one job

We split the account so that finding a winner and scaling a winner never competed for the same budget.

Naming followed one pattern, MARKET_CONCEPT_FORMAT_HOOK_V01, so anyone on the team could read a report and know exactly what was running.

06TestingA test has to earn its next euro

Before a concept went live, three things were written down: the maximum spend, the target cost per purchase and the date we'd look at it, allowing for purchase lag. At review, every concept got one of three verdicts.

Swipe the table sideways to see every column →

Over the 60-day rebuild: eight weekly launch cohorts, we tested 24 concepts. 8 reached the scaling campaign. That's a hit rate of 33%, in this test programme. Most tested concepts did not qualify for scaling. The system exists so the ones that do get found quickly and cheaply.

Clicks and thumb-stop rate helped explain a result. They never decided one. Only purchases at an acceptable cost did.

07ScalingScaling, step by step

We didn't scale on a calendar. Each budget level had to be earned, and each step was checked against the business as well as the ads: stock, fulfilment, creative supply and cash on the balance.

1. Week 1–2 · €8K/day

   Baseline on the new setup

   Spend held at €8K/day for 14 days. Baseline CPA €43.48 improved to €31.50 by the end of the hold; delivery interruptions stopped. Comparison KPIs use the separate 30-day windows.

2. Week 3 · €9.6K/day

   First step up

   First move from €8K to €9.6K (+20%) after three matured daily cohorts at CPA ≤€32. Further steps were at most 20%, at least 72 hours apart, with stock and funding checked first.

3. Week 5 · €19.9K/day

   The dip

   At €19.9K/day CPA rose from €31 to €36 for two days. Budget was held for six days while an overused concept was rotated. CPA returned to €31.80 before increases resumed.

4. Week 8–9 · €45K/day

   Holding the new level

   €45K/day reached within 60 days, then maintained throughout the after window. Three new concepts per week, four weeks of stock and a seven-day funding buffer supported delivery.

Top-ups were planned alongside each budget step, so funding was never the thing that decided when we could move.

When a step disappointed, we didn't rush to change creative. We checked in this order: tracking, stock, offer mix, landing page, then the ads. Most of the time, the cause sat outside Ads Manager.

08ResultsResults

Profit is reported after media spend, product and operating costs, creative production, our fees and the allocated account setup cost. ROAS, CPM and CPA use media spend only. Profit increased by €1.467M across the 30-day comparison windows, from €103K to €1.57M.

The brand now plans its next budget level around stock and creative.

Scaling became a scheduled operating process, with creative, stock and funding ready before each increase.

09TakeawaysWhat to take from this

• Know what a customer may cost. Agree it on first-order economics, then judge every test against it.
• Test buying reasons, not variations. Three distinct concepts teach you more than thirty versions of one.
• Get the setup ready before the budget. Spending limits and a single account cap growth earlier than most teams expect.
• Give every campaign one job. Testing finds winners, scaling feeds them, and neither steals from the other.
• Scale through gates. Stock, fulfilment, creative supply and funding have to be ready for the next level before the ads are.
• Put compliance before launch. In supplements, that's cheaper than any appeal.