Or in biopharma, where selling a product requires 10 years and $200M of r&d. The model is to sell your company or go public long before that point

In this case, your "customers" are venture funds and pharma companies, and your MVP is a well thought out experimental plan. You get more customer / investors by doing the experiments and using that data to design another experiment

They tried adapting the lean startup to life science with mixed success. For biopharma as a specific life sci sector other models work better than lean