Time to park it
When the destination is unknown, why prioritise the wrong metrics?
If you plug ‘Boston to New York’ into Apple Maps, it will give you route options, bearing in mind known knowns (available roads), known unknowns (typical traffic level variability) and more… You can choose on the basis of cost, difficulty of the road and more. Options are good, and Apple Maps quite reasonably doesn’t choose a single route for you.
If you didn’t say ‘New York’, however, the calculation just got a whole lot more interesting.
If, for example, the question was ‘where, within 4 hours, can I get the best dinner?’ or ‘the best route to get to New York and back without charging my Tesla?’, the same cast of known knowns would yield different answers, which might yield more options, depending on who you are. Handily, when I lived in San Francisco, asking for cycle routes would bear hill gradients in mind as well as traveled distance (hills are not my friend).
This tracks almost directly to our philosophy on path-to-market in pharma (and why it differs from ‘go to market’). The destination (the ‘market’) cannot possibly be fixed when development starts. It might be a good guess, but the value of that market is a guess that is likely to be off by orders of magnitude in early phase (unfortunately off in both directions). And, given the value of the market must be a key component in your choice of market, the question for Apple Maps is much more like our second type: ‘find me a destination for a drug like mine where I can be first to market, and gain share versus older competitors’.
‘Find me a market’ paths would obviously include multiple options of market, with time to market, cost to market and the value of the market if you make it.
Even if only one market destination is picked, there are still a variety of routes available - lower technical risk, greater cost to generate real value endpoints, etc. But, that variety is just more ‘known unknown’ for the calculation of viable launch options.
Leaving commercial value evaluations until the destination is fixed limits the options drastically. Limiting the launch options to a small subset chosen by R&D drastically limits the possibility to find a more viable commercial destination.