Max Hodak: What Really Kills Deep Tech Startups?
In a Nutshell
Deep tech startups fail due to broken infrastructure—procurement, hiring, performance management—not technology. Speed compounds through systems that eliminate approval bottlenecks, track true experimental costs, and enable distributed decision-making via continuous feedback rather than annual reviews. Success requires founders to develop independent judgment through hands-on experience, build custom internal tools as company operating systems, and focus relentlessly on reaching revenue sustainability rather than perpetual fundraising.
These notes were generated by AI and may contain inaccuracies.
Max Hodak is CEO of Science, a company developing infrastructure for deep tech startups. He spent nearly 20 years working on brain-computer interfaces, beginning his career as an undergraduate at Duke working with electrode implants in monkeys to study neural activity during joystick-based games. Science's main product is a retinal prosthesis: a chip implanted under the retina that restores vision to patients who have lost rods and cones. The device uses solar cells on the implant that receive infrared projections from camera-equipped glasses, creating electric fields to stimulate the retina directly. The product completed major clinical trials and was featured on the BBC and Time magazine, with one patient completing a 300-page novel using the device.
Picasso noted that art critics discuss form, structure, and meaning while artists discuss where to buy cheap turpentine. This illustrates the principle that amateurs talk strategy while professionals talk logistics. For deep tech companies, there are few universally applicable lessons because startups face continuous streams of daily facts requiring local decisions. However, certain topics recur consistently across deep tech companies, particularly around procurement, hiring, and performance management.
Pure software companies avoid physical procurement, which explains why VCs favor software investments. However, any company working with physical products purchases thousands of items continuously—computers, microscopes, electronics, 3D printers, resins, and PCBs. Credit cards work for founders, but the 17th employee needs purchasing capability. Handing out credit cards creates approval bottlenecks where founders question $3,000 power supply purchases, potentially delaying work for highly compensated employees. The cost of delay far exceeds potential savings from auctions.
Sign in to read the full notes
Get access to AI-generated notes, topic timestamps, and more.