<b>Case #019: The blended number that hid a bleeding account</b>

<b>Case #019: The blended number that hid a bleeding account</b>

We managed four campaigns under one offer and reported on the blend, because the blend looked great. Blended ROI: 44%. Healthy. We scaled the whole portfolio proportionally. That decision quietly poured money into a wound.

When we finally broke the blend apart, the truth was ugly:

— Campaign 1: $300/day, 95% ROI, the engine
— Campaign 2: $200/day, 30% ROI, fine
— Campaign 3: $150/day, 8% ROI, marginal
— Campaign 4: $250/day, -35% ROI, hemorrhaging

The 44% blend was Campaign 1 carrying three passengers, one of which was actively losing $87/day. By scaling the portfolio "proportionally," we'd been increasing budget on Campaign 4 in lockstep with our winner — feeding the bleed at the same rate we fed the engine.

The blended number wasn't just uninformative. It was actively hiding the decision we most needed to make.

— Killed Campaign 4 entirely
— Redirected its $250/day into Campaign 1 in 20% steps
— Left 2 and 3 flat to monitor

The portfolio's real ROI jumped not because anything improved, but because we stopped subsidizing a loser with a winner's profit.

— Pre-fix portfolio: $900/day, 44% blended ROI, ~$396/day profit
— Post-fix: $750/day, 78% blended ROI, ~$585/day profit

We spent less and made nearly $200/day more. The only thing that changed was that we stopped averaging away the truth.

The lesson: a blended ROI is where losing campaigns hide behind winning ones — judge every campaign on its own line, because a healthy average can contain an account that's quietly bleeding out.