# What is the minimum viable PARPA organization? # What is the minimum viable PARPA organization Perhaps one way to go about dissecting this question is to try to break it down into a series of characteristics. The intuition around the question “what would make a minimum viable institution?” is a nebulous metaphysical sense that the organization is now a Thing instead of just a collection of other Things. At some point a mixture of chocolate and peanut butter is no longer just a mixture of chocolate and peanut butter but a Reeses Peanut Butter cup as you give it more structure. Something something essential characteristics, something something reductiveness. The point is that there are not clear lines, a healthy dash of “you know it when you see it”, and several non-overlapping ways that the organization can be a Thing. I’ll try to legibly work my way through them to the best of my ability. Right now even if I hung up a (internet) shingle and said PARPA is open for business, it would not count as a Minimal Viable Organization. So labels (and even what other people think) don’t matter. As with any nebulous thing the best I can do is try to list a bunch of characteristics that more and more strongly suggest that there is indeed an MVI, the more there are in combination. And for good measure I’ll list a bunch of counter-characteristics that would suggest that there is no MVI **Number of people** - Does the organization need to be more than just me? No - there are one person VC firms and solo founder startups are a thing. While a minimum viable institution could be just one person, that wouldn’t be a stable state for the organization because you need more than one person to run enough programs to demonstrate viability. If [[An ARPA riff will need at least seven programs to demonstrate viability]], a program manager can run ~2 programs at a time[^1] and a program takes ~5 years, one person will take 10-20 years to do all seven programs. I’m ok with going slow but that’s a bit too slow. Add to the fact that the ~director will need to do other things like fundraise, administration etc. and it’s pretty clear that one person without the ability to bring on more people would not cut it as a minimum viable institution. The ability to bring on more people hinges primarily on the organization’s ability to pay them and its ability to deploy capital effectively. These two factors feel a bit like a Venn-diagram. Obviously, some effectively deployed money *would* go towards bringing on other PMs. But at the same time there is a scenario where you could bring on PMs but still be unable to deploy capital effectively. So that’s insufficient. **Ability to deploy a certain amount of capital effectively** The ability to deploy capital effectively is far more important than the amount of capital at the organization’s disposal. A classic trap is pouring capital on an organization that cannot use it effectively - I see Quibi and Magic Leap as two to-avoid examples here. Ability to deploy capital effectively is itself a nebulous status - while it *is* the ability to answer affirmatively to the question “If someone wrote you a $N check today, would you know what to do with it?” It *is not* an itemized list of expenses. Here the definition bifurcates a bit - the more agency other people in the organization In part this is going to be a gut call - how much money would I be comfortable taking? This comfort stems both from confidence that I won’t waste it, but enough confidence in the trajectory that I’ve laid out that the money itself won’t exert pressure on that trajectory. The more money you have, the stronger a pressure it will exert on your trajectory because of caring thresholds. [[People have different and unknown caring money thresholds]]. Thinking about it more, the ability to deploy capital effectively looks like some combination of having people lined up who I suspect would be PMs if the funding were available and a set of seedling projects that I would want to run as derisking moves for a program. There’s also the question of “how much capital” would you need to feel you could deploy effectively? My gut says “enough that I wouldn’t feel comfortable spending it all myself.” Which is effectively a large seed round’s worth (~$5m?). It is slightly larger than the standard Trust-building budget [[Trust-building timescale and budgets tend to be similar in different domains]] **A set of prioritized hypotheses** A big reason to create the organization in the first place is to test hypotheses that can only be tested in an organizational context. [[There are a number of experiments that seem like they can only be done in the context of an organization]]. So it’s important that a minimum viable organization has a a clear set of prioritized hypotheses that the organization is setting out to test. You could think of this as a series of concentric deriskings. The prioritized hypotheses also inform the amount of capital that can be effectively deployed. While the prioritized hypotheses are necessary for being able to deploy capital effectively, they are not sufficient. The ability to deploy capital effectively requires a level of detail (like who you would give the money to) that even if wrong or changed would not disprove the hypotheses. However, the ability to deploy capital effectively doesn’t necessarily mean you have a set of prioritized hypotheses - there could actually be a clearly effective step 1 that doesn’t have a bigger step two after it. **Skin in the Game** Other people’s opinions matter. There’s something to the difference between a project and an organization tied up in the idea that more than one person cares deeply about it. Maybe more than caring but that more than one person has skin-in-the-game in it’s success. This actually seems important. One of the classic startup founder moves is to say ‘we’ The trick is that there are actually many ways to get this skin-in-the-game. Getting funding is one way to do it - then the funder has some skin-in-the-game. Interestingly, for a funder to have SITG you need to be *above* their caring threshold, which means that there will be trajectory pressure, which forces you to have built up trajectory confidence. Having an employee is another way to generate SITG within other people. That, of course, requires funding. A partner is almost by definition, someone with SITG to the idea, regardless of salary. It might actually be possible to have a SITG from an idea? 🤔 It feels like having two program plans whose first steps were ‘shovel ready’ (effective capital deployment again!) would be right at the line of organization. But perhaps that is the SITG of people who would be unable to do a specific set of work otherwise? ### Conclusion A Minimum Effective Organization has roughly three components: other people with skin in the game, a set of prioritized hypotheses, and confidence about the ability to effectively deploy ~a large seed round’s worth of capital. [^1]: DARPA program managers often do more but they are also not trying to figure out a new institutional structure at the same time. [Web URL for this note](http://notes.benjaminreinhardt.com/What+is+the+minimum+viable+PARPA+organization) [Comment on this note](http://via.hypothes.is/http://notes.benjaminreinhardt.com/What+is+the+minimum+viable+PARPA+organization)