Case #015: The bid that was too high to win

Case #015: The bid that was too high to win

A mobile game install campaign was stuck. $2,900 spent, $2,100 back, a 28% loss that wouldn't move no matter what creative we threw at it. The instinct in a losing campaign is to bid up for better traffic. We did the opposite, on purpose.

The campaign was bidding $0.85 per click, chasing premium placements. The logic seemed sound — better inventory, better users. But the math underneath was broken: at $0.85 a click and a 2% install rate, our cost per install was $42 against a $30 payout. We were buying expensive traffic that couldn't possibly pay back.

We dropped the bid to $0.38 and accepted worse placements, betting volume and cheaper clicks would beat quality and price.

The lower bid changed the unit economics entirely:

— cost per click fell to $0.36 actual
— install rate dropped slightly to 1.7%, as expected
— but cost per install fell to $21 against the $30 payout

The turned-around campaign over ten days:

— $7,800 spent
— $11,400 revenue
— 46% ROI, up from a 28% loss

Worse traffic at the right price beat better traffic at the wrong one.

The lesson: in CPC buying, the most profitable user is rarely the most expensive one — sometimes you win by bidding to lose the auction for premium clicks.