One promo calendar, rebuilt. $2M in found revenue.
From the reworked calendar
Same budget, better timing + depth
Every code, window, and margin
When to run, how deep to go
The problem.
The brand ran promotions on instinct. Timing was a gut call; discount depth was whatever felt right. There was no baseline for what a "good" promo looked like, so no one could say which ones worked. Some promotions quietly cannibalized full-price demand. Others under- discounted and left money on the table. Same calendar every year, same guesses.
The approach.
I pulled every promotion out of the order history, dates, depth, codes, revenue, and margin, and built a baseline model: expected lift by discount tier, by calendar window, net of the sales the promo would have won anyway. Then I scored each historical promo against it, separating the winners from the ones that only looked good. The pattern became a simple playbook: when to run, how deep to go, and which windows to protect at full price.
The outcome.
Promo effectiveness rose 35% on the same budget, roughly $2M in additional promotional revenue over the following cycle. The team stopped guessing discount depth and started pricing every promo against a benchmark. The birthday sale, the seasonal drops, the flash windows, all of it now runs against a number instead of a hunch.
If your discount calendar is instinct and your team can't say which promo actually worked, I'll build you the baseline. One slot open for Q3 2026.