Deep-tech ventures fail not because any single barrier is insurmountable, but because the barriers are coupled — and must be cleared together.
The conventional account of the “valley of death” treats it as a funding gap: a shortage of capital between research grants and commercial revenue. That framing is too thin. In practice, a deep-tech venture faces at least four barriers — feedstock, capital, certification, and first-of-a-kind deployment — and the decisive difficulty is that they are coupled. You cannot secure the feedstock contract without the deployment commitment; you cannot get the deployment commitment without the certification; you cannot afford the certification without the capital; and you cannot raise the capital without the feedstock economics. Each barrier is a precondition for clearing the others.
The MASH experience is a clean illustration. The cashew-residue feedstock only had value once a fuel specification (certification) made the output tradable; the certification only mattered once a shipowner (deployment) agreed to a trial; the trial only happened because a blended funding stack (capital) had de-risked the chemistry far enough to attempt it. Pull any thread and the whole knot loosens.
This reframing has a practical consequence. It suggests that the right unit of strategy is not the individual barrier but the sequence of coupled moves that clears them in a mutually reinforcing order — and that the founder’s core skill is choreographing that sequence, not optimising any single step.
It is a claim that can be tested against other deep-tech histories.
The open question
Can the coupling structure of deep-tech barriers be characterised generally enough to predict which ventures will stall, and where?