I intellectually really like these experiments, but one of the things virtually every startup advisor tells you is "only innovate when and where you must" -- basic things like how you do accounting, offices, etc. should be standard, so you don't screw them up, so you don't think about them too much, etc.
This is how a financier would think about it (and I'm not saying it is always wrong). There's some finite risk allowance, and the idea is that you should load up 100% on business model risk-- which means no experimenting with open allocation, remote work, etc.-- because business model risk is what the VCs are good at analyzing and there's a bit of golden hammer syndrome ("when you have a great hammer, everything looks like a nail.")
As a practical directive, I think it's dead fucking wrong (practically; morally, people can run their businesses how they wish) and here's why. If you load up 100% on business model risk, then 0% gets allocated to employees. They get no autonomy, work long hours, and generally get abused because "we don't have any slack in the schedule". Well? If your business model is innovative but the only thing you're delivering to employees is the stereotypical soul-sucking corporate job, you're going to have to give 25-35% annual raises each year (as Wall Street does) to keep anyone good. That's expensive, unsustainable, and by necessity leads to an up-or-out culture that fries the place anyway. Of course, startups replace this with equity options and hope the exponential curve of the valuation will deliver the implicit raises, but anyone who can do basic math on a typical engineer grant (typically $5-20k per year at-valuation) is going to find it weak as a motivational tool.
If people have the autonomy to direct their own careers and work on things that interest them, they're a lot less expensive in the long run, because it doesn't take a 30% pay bump and a newer, shinier management title each year to keep them on. This is even more true in startups. People expect, e.g., Microsoft or Goldman, to be bureaucratic. If people go into a startup and find that the traditional hierarchy and closed-allocation regime are in place, they're going to feel cheated and get pissed.
If you want good people and you want them to be motivated by the work instead of an exponentially growing paycheck, that means you have to experiment and give them unusually high degrees of autonomy.
I think we all agree startups should innovate a lot on their core product where it can be a useful differentiator. I think this goes far beyond "business model" -- things like which programming language to use for the core product are good things to innovate on. Using untested/unpopular technologies like Linux, functional languages (well, really just interpreted languages, functional was still a small subset), Apache, etc. became useful advantages for some companies in ~2000. Using the cloud in 2008-2010, or adopting social and then mobile from 2008-2012. Not sure what it is today (I'd like to think it's hostproof/security-from-NSA, but that's probably not it for most businesses.)
I hope we all agree startups have finite resources and shouldn't try to innovate (or really even spend much time on) the really mundane things which don't matter to that particular startup. Legal paperwork, trash disposal, etc.
It's an interesting question which things are in the middle, and how much innovation/experimentation is reasonable there. Hiring is clearly core, so things around hiring might be more legitimate grounds for experimentation than others. I don't think cash compensation is the pain point for most of that, though -- how people pick projects, how people are recruited, etc.
This is how a financier would think about it (and I'm not saying it is always wrong). There's some finite risk allowance, and the idea is that you should load up 100% on business model risk-- which means no experimenting with open allocation, remote work, etc.-- because business model risk is what the VCs are good at analyzing and there's a bit of golden hammer syndrome ("when you have a great hammer, everything looks like a nail.")
As a practical directive, I think it's dead fucking wrong (practically; morally, people can run their businesses how they wish) and here's why. If you load up 100% on business model risk, then 0% gets allocated to employees. They get no autonomy, work long hours, and generally get abused because "we don't have any slack in the schedule". Well? If your business model is innovative but the only thing you're delivering to employees is the stereotypical soul-sucking corporate job, you're going to have to give 25-35% annual raises each year (as Wall Street does) to keep anyone good. That's expensive, unsustainable, and by necessity leads to an up-or-out culture that fries the place anyway. Of course, startups replace this with equity options and hope the exponential curve of the valuation will deliver the implicit raises, but anyone who can do basic math on a typical engineer grant (typically $5-20k per year at-valuation) is going to find it weak as a motivational tool.
If people have the autonomy to direct their own careers and work on things that interest them, they're a lot less expensive in the long run, because it doesn't take a 30% pay bump and a newer, shinier management title each year to keep them on. This is even more true in startups. People expect, e.g., Microsoft or Goldman, to be bureaucratic. If people go into a startup and find that the traditional hierarchy and closed-allocation regime are in place, they're going to feel cheated and get pissed.
If you want good people and you want them to be motivated by the work instead of an exponentially growing paycheck, that means you have to experiment and give them unusually high degrees of autonomy.