Materials Science Startup Lessons

I’ve learned a lot over the last year of running polySpectra, as part of the Cyclotron Road accelerator program at LBNL. Here are a few bite-sized tips that might help you in your endeavors:

  • A prototype is worth a million words. Most people have a very hard time hearing descriptions of chemical and physical properties and translating that into an accurate mental image. Even people with PhDs in your field may not immediately understand what you are talking about (although they will pretend they do, which is even worse). If a picture is worth a thousand, putting a prototype in their hands is worth a million.

  • Never guess someone else’s margins. Don’t try to guess what another firm’s profit margins are on a given chemical, material or product. You have no idea, it could be 2% or 2000%, and it will take a lot of digging and a lot of trust-building in the space before you’ll be able to find out. The bigger the firm, the more meaningless the margin is in terms of competitive advantage - they’ll find another way to balance the books if they need to cut you out.

  • Science is hard. When you’ve written that awesome paper and filed your patents, and you feel like you’re 90% of the way to a commercially viable product - guess again. The brutal reality is that you’re more like 10% of the way there. A Caltech professor once told me, “to have a paper it has to work twice, to have a product it has to work two thousand times.” Stick with it though. Focus on reliability and reproducibility (sadly these are two areas in which your academic training is most likely useless, and possibly detrimental). Novelty might get you citations, reliability is what gets you customers.

  • Stay lean longer. Physical science innovations are expensive and time-consuming to bring to market. The hardest technical problems cannot be solved simply by throwing a bunch of money at the problem. There are lots of ways that you can think entrepreneurially and develop your technology with a market-aware mindset without actually spinning out a company (yet). Another Caltech professor told me, “I realized lately that we’ve been spinning these things out too early. They probably would have had a better chance if we’d cooked them in the lab for another year.”