Francis Ruan

Index / Work

Suntag Co.

FounderMay 2025 to present$5,000 pre-seed300+ waitlistLos Angeles

You have to name a price before anyone has ever paid you one. Tooling gets committed first. That is the whole problem, and I got a piece of it wrong.

The problem

Hardware pricing is set before you have demand data, because tooling and the first production order are committed in advance. Price too low and you delete the margin that funds the second run. Price too high and you hold inventory with no signal about why.

What I owned

All of it. Solar viability testing, the interface in Figma, the prototype code, and the demos I ran to watch people react. There was nobody to hand the pricing question to.

Approach

  • Tested solar output first, because the ceiling capped what the product could honestly claim.
  • Treated demos and the waitlist as directional signal, never as stated willingness to pay.
  • Reduced it to three inputs: landed unit cost, tolerable price, minimum viable order.

Result

Funded pre-launch on a $5,000 pre-seed with a waitlist above 300. That is validation, not revenue, and I will not describe it as more.

The pricing decision, split into what was measured and what was assumed.
InputWhat I knewWhat I assumed
Solar outputMeasured in testingReal use underperforms the bench
Demand300+ signups at zero costSome fraction converts at some price
Landed costQuotes at several order sizesFirst run holds quoted tolerances
Price toleranceVerbal reactions from demosInterest survives a real checkout
Order sizeSupplier tooling and minimumsRun one funds run two

Every row on the right is an assumption a buyer would test. I have not published unit costs or margins, because the numbers are not stable enough to defend.

What I got wrong

I treated the waitlist as demand. A waitlist measures interest at zero cost and says nothing about the price at which that interest survives.

Running it again I would take refundable deposits at two price points before committing to tooling. That converts a soft signal into a number I could actually defend, and it costs almost nothing.

If any of this is worth an argument, I would like to have it. fyruan@usc.edu