Case #035: Buying November in September

Case #035: Buying November in September

A gifting ecommerce offer, $4,500 budget. The instinct with seasonal products is to go heavy in peak season, and everyone does — which is exactly why peak season is the worst time to buy. In late November my target CPMs run double, sometimes triple, because every buyer on earth is bidding for the same holiday eyeballs.

So I ran the campaign backwards. I started buying in September, at a loss on purpose, to build a retargeting and lookalike pool while inventory was cheap.

The September arc was ugly by design: $1,600 spent, $980 back. Minus 39%, and I logged it without flinching because the real asset wasn't the revenue — it was the 9,400-person pixel pool of people who'd browsed the product at off-season CPMs of $4 instead of December's $11.

Then I held. In the last week of November I activated lookalikes seeded from that warm pool. The seeded audiences converted at a $19 CPA against the $31 cold buyers were paying, because my seed was built from real intent gathered cheaply months earlier.

Full arc, September buy plus the November activation, measured at season close: $4,500 spent, $7,830 back. 74% ROI — but the profit all landed in one frantic November week, funded by a September loss I'd planned in advance.

The lesson: in a seasonal vertical, the cheapest time to buy your customers is months before you need them — eat the off-season loss to own a warm pool when everyone else is paying triple.