The question every baker asks before every drop: how many? The honest answer is that nobody knows — including the bakeries that look like they know. What they have isn't a crystal ball; it's three or four drops' worth of data and the habit of reading it.
Your first three drops are experiments wearing an apron. Here's what each one is telling you.
Drop one: the baseline
Whatever you sell in drop one is your baseline, and the only real mistake is baking so much you can't sell out. (We've said it before: start smaller than feels right. "Sold out" is marketing; leftovers are just sad math.)
The number to write down isn't just total sold — it's how fast. Sold out in ten minutes, two hours, or never is three different businesses.
Drop two: the shape
Run the same menu again. Same items, maybe 20 to 30 percent more of whatever vanished first. Now you learn the shape of your demand: does the same crowd come back? Do the morning buns always die first while the shortbread lingers?
Drop three: the lever
By drop three you know your baseline and your shape, so you get to pull one lever. One, not three — change everything at once and you learn nothing.
- Sold out in minutes twice? The lever is price. Raise it 10–15% and watch what happens. (Usually: nothing, except you earn more.)
- One item always dies, one always lingers? The lever is menu. Give the lingerer's oven space to the star, or make the star's variant the special.
- Comfortable sellouts, happy pace? The lever is quantity. Step it up 25% and see if the demand keeps stretching.
The waitlist is free market research
When an item sells out on OurCakewalk, customers can leave their number for a restock text. That waitlist count is the purest demand signal you'll ever get — it's people who tried to give you money and couldn't. Three people waitlisted for macarons is not a rounding error; it's twelve more macarons next week.
When the answer is "don't grow"
One more honest thing: some of the best home bakeries plateau on purpose. If 30 boxes a week fits your life and sells out every time, that's not a business failing to scale — that's a business that works. The data tells you what you could do. Your Saturday mornings get a vote on what you should.