Fund the Company Now. Price It When the Evidence Arrives.

A deep-tech founder is often asked to price an experiment before raising enough money to run it.

The founder believes the science will work and argues for the value that success could create. The investor sees the same possibility but has to price the risk that the decisive result may fail. Both are being rational. The evidence that could settle the argument does not exist yet.

Today, that disagreement usually ends in one of three ways. The company accepts a low valuation, the investor accepts a high one, or the money is divided into tranches and released as milestones are met. The first two force one side to make the wrong bet at signing. Tranches create a different problem. They hold back the money the company needs to produce the proof.

I built PROVE around a simpler idea: fund the company now and let verified evidence determine the price later.

PROVE stands for Price Resolved On Verified Evidence. The investor pays the full amount at signing. Before the key experiment begins, the company and investor agree on one technical readout, the method used to measure it, the thresholds, the deadline, and a table of valuation caps. An independent Referee receives the protocol before the data exist and later confirms the result. That result selects the valuation cap at which the investment converts.

The money is never contingent. Only the price is.

Consider a $3 million investment. The parties might agree on a $60 million cap if the experiment is a Hit, a $35 million cap for a Partial result, and a $20 million cap for a Miss. At conversion, those caps would give the investor 5 percent, 8.6 percent, or 15 percent ownership before new money. Everyone can see the full range on the day of signing. Nobody can rewrite the experiment after seeing the data, and the investor does not get to decide alone whether the result counts.

This is more useful than negotiating harder because the disagreement is scientific, not rhetorical. A founder should earn the higher price by producing the result. An investor should receive more ownership if the central technical claim fails. Until then, the company should have the capital and freedom to do the work.

PROVE is designed for AI, biotech, and deep-tech companies whose value depends heavily on a result that can be measured, pre-registered, and independently verified within a financing cycle. That could be performance on a sealed AI test set, a blinded diagnostic study, manufacturing yield, battery performance, or a laboratory-confirmed hit rate for AI-designed molecules.

It is the wrong instrument when the real uncertainty is customer demand, founder execution, or a business model that will take years to reveal itself. PROVE does not replace diligence, and it does not manufacture certainty where no decisive experiment exists. It separates one specific risk from the rest of the company and gives that risk a price.

The standard is open because a financing mechanism becomes useful only when founders, investors, lawyers, laboratories, and future funders can recognize the same rules. The full PROVE Standard includes the agreement, a readout library, worked ownership examples, and safeguards against manipulation. It is licensed under Creative Commons Attribution 4.0 so others can test it, adapt it, and improve it.

We have spent years making scientific progress easier to measure. The financing should finally learn how to listen.