Every account we've ever taken over from another agency has the same story: spend went up because the account felt like it was working, not because the data said to push it. A few good days. A founder's gut. A screenshot of one ad that "popped." None of that is a scaling decision — it's a guess with a budget attached.
Structured testing, not vibes
Before we scale anything, we want a proven pattern, not a proven moment. That means running new creative and audience combinations through a defined testing window — long enough for Meta's delivery system to exit the learning phase and for results to stabilize, not just long enough to see one good day.
We're looking for consistency across that window, not a single spike. A winner that holds for a week under stable conditions tells us something. A winner that spiked for six hours doesn't.
What we actually check before scaling
- Sample size — enough spend and enough conversions that the numbers aren't noise.
- Stability — cost-per-result holding steady across the testing window, not swinging wildly day to day.
- Frequency — the audience isn't already fatigued on the creative we're about to pour more budget behind.
- Downstream data — the "win" holds up in actual sales or qualified leads, not just cheap top-funnel clicks.
Why this matters more than the ad itself
An agency that scales on vibes isn't taking on more risk than you are — they're taking on none, because it's your budget. Disciplined testing is slower to feel exciting week to week, but it's the difference between an account that compounds and one that lurches between hot streaks and hangovers.
This is also why we don't publish "case study" percentages we can't stand behind. If a number goes on a page with our name on it, we can walk you through exactly how it was tested.