Case #005: The turnaround hiding under a bid floor
The campaign was dead on arrival. A push-traffic install offer, $0.42 payout, and we couldn't get below $0.55 cost-per-install no matter how we sliced the creative. Three days, $310 spent, $237 back. A clean loss, ready for the graveyard.
Before killing it, we pulled the placement report. One line jumped out: 71% of our spend was going to a single source ID that converted at $0.71 per install. The other 29% of spend, spread across smaller sources, was converting at $0.38.
We'd been judging the whole campaign by the average. The average was a lie told by one bad source drowning out a profitable one.
— Blacklisted the whale source eating 71% of budget
— Re-aimed the entire bid at the cheap-converting tail
— Cost-per-install dropped from $0.55 to $0.39 in 48 hours
The turnaround:
— Days 1-3 (pre-fix): $310 spent, $237 revenue, -24% ROI
— Days 4-10 (post-fix): $1,460 spent, $1,840 revenue, +26% ROI
It was never a bad campaign. It was a good campaign with one expensive source standing in front of it. Source-level blacklisting on day 1 instead of day 4 would have saved us the entire loss.
The lesson: a campaign average is the most dangerous number on your dashboard — one fat source can disguise a winner as a corpse, so read the placement report before you call time of death.
The Green Day
@greenday_roi
Case #005: The turnaround hiding under a bid floor
Этот пост опубликован в Telegram-канале The Green Day. Подписаться можно по ссылке: @greenday_roi.