Otterbein's
Cookies
A team that didn't think ecommerce worked for cookies. Revenue is up 3.7x, discounting is down two thirds, and they're tripling paid media spend.
Selling cookies the hard way.
No paid media. No conversion infrastructure. And discounts were eating 34.6% of gross sales — a third of the topline handed back at checkout, because price was the only lever anyone had.
The team was skeptical that ecommerce could work for the category at all. Cookies are low AOV, impulse, and expensive to ship. They didn't think the math would close.
We don't start with ads. We start with whether the store can hold them.
Phase one rebuilt the commercial foundation: pricing architecture, bundle and offer structure, and a P&L built around AOV and LTV rather than volume. On top of that, a conversion-optimized store, retention flows for welcome, abandoned cart and reviews, subscription setup, and full analytics instrumentation.
The brand stopped buying its own revenue back at checkout. That single change is worth more at scale than any ROAS improvement, because it moves every future order.
We cut spend roughly in half. The business kept 36% more.
Month one ran hot and uneven and bought volume at 2.42x with CAC 24% over target. That is the expected shape of a first month — the job was to generate enough real purchase data to rebuild against.
Month two is where the system started working. Weekly creative briefing, audience testing, and budget reallocation against actual purchase data rather than platform defaults.
The traffic got better, not just cheaper. Impressions fell 44.5%, but click-through rose from 4.5% to 6.1% and click-to-purchase rose from 1.71% to 2.57%. Fewer people saw the ads. More of the right ones did.
They're tripling paid media spend.
Revenue has grown 3.7x since April. We came in at the end of May. Paid media went from nothing to a channel returning 3.46x with CAC 16% under target, discounting is down from 34.6% of gross sales to 12.1%, and refunds for the most recent month were zero.
Month two was deliberately constrained while we rebuilt the account. Once the unit economics cleared target, the budget stopped being a safety measure and became the constraint.