The 10x Deep-Tech Screen
Back breakthroughs large enough to survive technical execution risk
- Difficulty
- Advanced
- Time to result
- ~months to results
- Steps
- 5
- Confidence
- 94%
The 10x Deep-Tech Screen evaluates whether a laboratory or garage breakthrough has enough advantage to become a venture-scale company. First establish the relevant state of the art, then look for roughly a tenfold improvement in a decisive metric such as cost, speed, or efficiency. A twofold improvement may leave too little margin to survive engineering, regulation, financing, and adoption. Next separate technical risk—whether the system can be built—from market risk—whether buyers want it. Holman favors technologies serving vast existing industrial markets, where demand is already evident. The pivotal milestone is a working deployment that removes core technical uncertainty. Once that proof exists, capital-intensive replication may use project finance or debt rather than repeatedly selling venture equity. The screen therefore pairs a high technical-risk tolerance with a strict requirement for exceptional upside and established demand.
Origin
Extracted from The Tim Ferriss Show as Holman explains how his fund selects technologies to move from a lab or garage into a startup.
Core principles
- 01A breakthrough needs enough advantage to absorb execution friction
- 02Deep tech exchanges market risk for technical risk
- 03A working technical demonstration collapses a major risk category
- 04Large established industries can supply demand without creating a new market
How to run it
- 1
Benchmark the incumbent
Define the current best solution and the metric customers or operators care about. Use a real industrial baseline rather than a convenient internal comparison.
Pro tip Choose cost, speed, efficiency, safety, or another metric tied to adoption.
- 2
Look for a tenfold window
Test whether the new technology can be roughly ten times better on at least one consequential metric. Treat smaller gains as vulnerable to implementation friction.
Pro tip A tenfold cost or efficiency shift can support an entirely different business model.
Watch out Do not round a modest improvement up to a breakthrough.
- 3
Separate the risks
Write down the remaining technical, market, regulatory, and execution risks independently. Do not import software venture assumptions into a physics-heavy company.
Pro tip Define the experiment that would retire the central technical risk.
Watch out A large market does not prove that the technology works.
- 4
Confirm existing demand
Identify the large industrial market already paying to solve the problem. Favor a breakthrough that replaces a costly incumbent over one that must invent both technology and demand.
Pro tip Measure the incumbent industry's actual spend.
Watch out A broad theoretical TAM is not the same as reachable industrial demand.
- 5
Plan the graduation milestone
Specify the working demonstration after which technical risk materially falls. Map the debt, project finance, or commercial structures that can fund replication after that point.
Pro tip Make the first functioning installation an explicit financing inflection point.
Watch out Do not assume repeated venture rounds are the only route to scale.
In the wild
Holman describes a small reactor designed for a mile-deep borehole, using surrounding rock for containment and water pressure from gravity for cooling. Its claimed advantage combines factory production with radically simpler containment.
→ A successful first deployment would visibly retire core technical risk in an existing, enormous energy market.
A crewless sailing ship targets an industry where Holman says most spending is burned as fuel. Smaller vessels would trade slower water speed for less port congestion and lower fuel requirements.
→ The concept targets a vast existing shipping market with a substantial cost and operating-model advantage.
Common mistakes
Backing incremental gains
A twofold improvement may not leave enough room to absorb the difficulty of commercializing deep technology.
Using software risk assumptions
MVP and product-market-fit thinking does not replace technical-risk milestones for reactors, ships, or new materials.
Financing every stage with equity
After technical proof, project finance and debt may be better suited to repeat deployments.
Is it for you?
Best for
It is best for evaluating pre-seed technologies aimed at large, established industries.
Not ideal for
It is not ideal for ordinary software products whose primary uncertainty is customer demand rather than technical feasibility.
From the transcript
“I usually get involved when I see a technology that I think is 10 times better than state-of-the-art”
“10 times cheaper, 10 times faster, 10 times more efficient, 10 times on any metric could be a good window.”
“Basically no market risk once we get through the technical risk.”
From the episode
#827: Pablos Holman — One of The Scariest Hackers I’ve Ever Met
Pablos Holman