5.5x more spend, at a higher return.

Stuck at $2K a month, and the signal was messy.

Treadlite was capped at $2,000 a month at a 1.9x return. Budget was leaking into low-intent queries, and that messy signal was feeding automated bidding the wrong data.

The goal was not a quick spike. It was more spend at a higher return, scaled in steps the account could actually hold.

Clean the signal, then scale in steps.

• Build real negative lists so budget stops leaking into low-intent queries.
• Rebuild structure around the highest-intent buyers so Smart Bidding learns from clean data.
• Scale budget in roughly 20% steps, only after each step holds its return.
• Judge every step on real return, not clicks or impressions.

Seven months, 5.5x more spend, 84% higher return.

This was a turnaround, not a launch. Seven months of tightening the signal and scaling only what held: $2,000 a month at 1.9x became $11,000 a month at 3.5x.

Across the run: spend up 5.5x, return up 84%, roughly 10x monthly revenue at the new spend level.