That said personalized medicine should end up being interesting, but I think it's really, really early days.
I think I see the experiments you're doing more as internal development. I.e. I run bits of my code dozens of times a day to make sure it's working (or not), which I see as analogous to running bench experiments.
Also, I'm thinking more about traditional pharma biotech where you can run tons of experiments but the drug has to be put into people at some point and there's actually a lot that can go wrong there that can't be easily forecast. The only time you get hints at whether your drug is "viable" is when you're in clinical trials and even then until succeeding in Phase III there's a ton of potentially catastrophic risk.
For the record I was a bench biochemist for the better part of a decade, so under no circumstances do I mean to denigrate bench science.
Internet-type startups, at their core, are about marketing more than they are about technology. Getting the technology right is absolutely no guarantee of success, but you can hack together some PHP and -- if you get the marketing right -- make a shitload of money.
Psychology is kinda hard to do rigorously, so you test marketing the only place you really can -- in the market.
Conversely, the question "does this material prevent infections?" can actually be studied, to a significant extent, in controlled lab experiments. You don't need the marketplace to answer that question.
And for many cases -- such as the OP's -- it's obvious that if the answer is "yes" then you have a valuable product. Period. End of story. Infection-resistant materials aren't a fucking social network. If they work, they are obviously and immediately useful. You'll need to get the business side right, but it's really less difficult than the science itself because hospitals are begging for products to buy that address this problem.
So the question is more "can we solve this totally obviously extant and well-defined and obviously significant problem" rather than "have we identified a problem whose solution will let us win some pocket change of a big number of consumers?"
Figuring out if the science works out IS the MVP test. If you have a product that works your customers will be lining up.
Consider the pharma route. Does it solve something that slots in, or will the process be hard, or maybe they have something close enough that disrupting the pipeline isn't worth it? Many projects don't get into testing, many testing projects don't get purchased, and as generally only one pharma will want to purchase, scale won't be hit (so you're dependent), and many purchased projects won't get acquired.
There are other routes and they have their own business risks. The MVP / market analysis process is different for biotech, but still should happen. It's pretty fascinating!
Right. I think we're all violently agreeing.
Biotech companies do need MVPs and market analysis. My observation is just:
1) The idea of testing product-market fit by building a complete product and putting it out there in uncontrolled settings doesn't really work for a lot of biotech companies. Especially anything that's going to need regulators' approval.
2) Compared with IT startups, Biotech companies are more likely to fail due to having the wrong answer for "but can we make it work?" rather than "is there a market?". So the MVPs optimize for testing former question, which usually means controlled experiments in a lab. There are counter-examples on both sides, I'm sure, just in general.
> Does it solve something that slots in
I'm not sure what you mean by this...
> they have something close enough that disrupting the pipeline isn't worth it
This is another incident of thinking that IT startups are the same as Biotech startups, or any science startup. Marginal increments in IT can be worth BILLIONS, marginal increments in Biotech aren't worth the paper they are printed on (that may be a bit of an exaggeration), and the biggest reason is cost to get into market. IT has low costs if any to enter a market place, except in the rare exceptions where the FCC is involved, the average cost to get a drug to market is something like $500MM and YEARS of regulatory approval. It's not worth investing in a marginal increment, you invest in moonshots and as you collect more data and better understand how the drug performs to either continue to invest (if it remains promising) or shelve it if it only looks marginally better.
While the market analysis needs to happen it is much less rigorous than IT startups. Hey 100,000 people die every year from infections acquired in hospitals, great lemme call around see what hospitals are willing to pay and sell to them. Lets not forget the other 600,000 that acquire infections from hospitals, for which under the Affordable Care Act the hospital has to pay for. Great sounds like I have a business as long as my product doesn't surpass a certain price threshold. In biotech you are either a billion dollars or a bust, there is no in-between because new products cost so much you have to have those high returns.
Not many IT startups NEED $500MM just to get their first sale.
Your product will not only require extensive marketing to make it into a major hospital system, it will also need to be cheaper than brasses and bronzes.