The most expensive mistake in physical products is inventory nobody wanted. The second most expensive is skipping a product people would have loved because you could not risk the order. Both come from the same gap: not knowing real demand. Here is how to close it.
Weak signals: likes, surveys, waitlists
Polls and comment sections measure politeness, not demand. Waitlist signups are better, but they are still free: a common rule of thumb is that only a fraction of a waitlist converts to a paid order, and you will not know which fraction until the money moment. If your production decision is expensive, free signals are not enough to bet on.
The strong signal: charged orders
A backer who has actually paid is demand you can take to a manufacturer. That is the entire logic of the all-or-nothing campaign: buyers are charged at checkout like a normal order, and if the goal is not reached by the deadline, every backer is automatically refunded. Buyers commit honestly because they are protected, and you learn the truth because they paid.
Make the test map to your real decision
- Set the goal at your production minimum, the smallest run where the economics work. Hitting it means the product is real. How to pick the number.
- Set a max cap at the most you can actually deliver, so a viral spike cannot oversell you.
- Keep the window short. 7 to 14 days concentrates your audience's attention and gives you an answer fast.
If it misses, you still won
A failed campaign costs you a bulk refund click and a little pride. Compare that to a garage of unsold stock. You also leave with real data: how many people put money down, at what price, from which channel. Adjust the product or the price and run it again. How refunds work.