There is a specific number where most fashion brands stop scaling on Meta. It is not a coincidence that it sits around €100K of monthly attributed revenue. It is the point at which the operational structure that got the brand to €100K becomes the exact thing preventing it from reaching €250K. The brand has four overlapping bottlenecks. Each one looks like the cause when in isolation. Solved in the wrong order, the brand burns spend trying to fix the symptom.

The plateau pattern

Month 1–4: revenue grows. Month 5–7: revenue plateaus. Month 8: founders panic, raise budget, see ROAS collapse. Month 9: budget cut back, revenue stabilizes at the plateau. Month 10–12: the brand convinces itself that "€100K is just what we are." It is not. It is the ceiling of the current operational architecture, not the ceiling of the brand.

Bottleneck one — Creative velocity

At €30K/month, two new creatives a month is enough. At €100K/month, two is the rate of decay. The audience consumes creative faster than the production pipeline can refresh it. The pipeline becomes the constraint. Founders who try to solve this by ordering "more UGC" without changing the brief, review, and retire infrastructure end up with more bad creative, not more good creative.

Bottleneck two — Audience saturation

Fashion brands typically launch on Meta with broad targeting plus 2–3 lookalikes that perform well early. By month four, those lookalikes have been served the same hero creatives so many times that frequency is above 5.0 and the algorithm starts widening on its own — pulling in cold audiences without the corresponding creative variation to convert them.

The fix is not new audiences. The fix is rebuilding audience architecture around buying intent signals — site behaviour, email engagement, returning visitors — and feeding fresh creative angles to each segment.

Bottleneck three — Attribution decay

Most brands set up their pixel and CAPI in year one, then never revisit. By year two, events have drifted, duplicate firings have crept in, and the data Meta is optimizing against is 15–20% noise. The algorithm continues working — but it is working on degraded signal, which means it is steering toward decisions that look correct in-platform and are wrong in the actual P&L.

Fixing attribution before scaling spend is non-negotiable. Spending more on broken signal does not produce more revenue — it produces more confusion.

“Spending more on broken signal does not produce more revenue. It produces more confusion.”

Bottleneck four — Margin compression

The final bottleneck is the one founders see most clearly — CPMs rise, AOV stagnates, and the unit economics tighten. The instinct is to push for higher ROAS, which is the wrong instinct. The fix is on the offer side: bundling, post-purchase upsells, subscription mechanics, and creative that justifies higher AOV. Meta cannot fix margin — only the offer can.

The order in which to dismantle

The correct sequence is:

  • 1. Attribution — until the data is true, no optimization is real.
  • 2. Creative velocity — install the pipeline before scaling spend.
  • 3. Audience architecture — rebuild around real intent signals.
  • 4. Margin — refine the offer to absorb the next scaling tranche.

Most agencies try to solve audience and creative first because those feel like the "ad agency" problems. They are downstream. The brands that break through the plateau are the ones who fix the foundation first — and have the patience to do it.