Here's an article in Drug Discovery Today on "virtual pharmaceutical companies", and people who've been around the industry for some years must be stifling yawns already. That idea has been around a long time. The authors here defined a "VPC" as one that has a small managerial core, and outsources almost everything else:
The goal of a VPC is to reach fast proof of concept (PoC) at modest cost, which is enabled by the lack of expensive corporate infrastructure to be used for the project and by foregoing activities, such as synthesis optimization, which are unnecessary for the demonstration of PoC. . .The term ‘virtual’ refers to the business model of such a company based on the managerial core, which coordinates all activities with external providers, and on the lack of internal production or development facilities, rather than to the usage of the internet or electronic communication. Any service provider available on the market can be chosen for a project, because almost no internal investments in fixed assets are made.
And by necessity, such a company lives only to make deals with a bigger (non-virtual) company, one that can actually do the clinical trials, manufacturing, regulatory, sales and so on. There's another necessity - such a company has to get pretty nice chemical matter pretty quickly, it seems to me, in order to have something to develop. The longer you go digging through different chemical series and funny-looking SAR, all while doing it with outsourced chemistry and biology, the worse off you're going to be. If things are straightforward, it could work - but when things are straightforward, a lot of stuff can work. The point of having your own scientists (well, one big point) is for them to be able to react in real time to data and make their own decisions on where to go next. The better outsourcing people can do some of that, too, but their costs are not that big a savings, for that very reason. And it's never going to be as nimble as having your own researchers in-house. (If your own people aren't any more nimble than lower-priced contract workers, you have a different problem).
The people actually doing the managing have to be rather competent, too:
All these points suggest that the know-how and abilities of the members of the core management team are central to the success of a VPC, because they are the only ones with the full in-depth knowledge concerning the project. The managers must have strong industrial and academic networks, be decisive and unafraid to pull the plug on unpromising projects. They further need extensive expertise in drug development and clinical trial conduction, proven leadership and project management skills, entrepreneurial spirit and proficiency in handling suppliers. Of course, the crucial dependency on the skills of every single team member leaves little room for mistakes or incompetency, and the survival of a VPC might be endangered if one of its core members resigns unexpectedly
I think that the authors wanted to say "incompetence" rather than "incompetency" up there, but I believe that they're all native German speakers, so no problem. If that had come from some US-based consultants, I would have put it down to the same mental habit that makes people say "utilized" instead of "used". But the point is a good one: the smaller the organization, the less room there is to hide. A really large company can hol (and indeed, tends to accumulate) plenty of people who need the cover.
The paper goes on to detail several different ways that a VPC can work with a larger company. One of the ones I'm most curious about is the example furnished by Chorus and Eli Lilly. Chorus was founded from within Lilly as a do-everything-by-outsourcing team, and over the yeras, Lilly's made a number of glowing statements about how well they've worked out. I have, of course, no inside knowledge on the subject, but at the same time, many other large companies seem to have passed on the opportunity to do the same thing.
I continue to see the "VPC" model as a real option, but only in special situations. When there's a leg up on the chemistry and/or biology (a program abandoned by a larger company for business reasons, an older compound repurposed), then I think it can work. But trying it completely from the ground up seems problematic to me, but that could be because I've always worked in companies with in-house research. And it's true that even the stuff that's going on right down the hall doesn't work out all that often. One response to that is to say "Well, then, why not do the same thing more cheaply?" But another response is "If the odds are bad with your own people under your own roof, what are they when you contract everything out?"