I’ve been thinking about peer-to-peer models today, and what early learnings signify for how these businesses might work at scale.

The truly peer-to-peer online learning platforms, like Udemy or Skillshare*, risk cross-pollination of their supply and demand. Someone who teaches a successful Udemy class might go on to post a class on Skillshare. And someone who enjoys taking classes on Udemy might also take classes on Skillshare.

No peer-to-peer online learning platform can own its supply or demand, which has made it difficult for a clear winner to emerge at scale. (Other platforms, like Coursera or Udacity, take a more curated approach to who can teach a class, perhaps for this reason.)

Similarly, as the battle for a winner in the ridesharing space has escalated, Uber and Lyft* have become indistinguishable at scale. Earlier on, Lyft was known for its “friend with a car” vibe, whereas Uber had a luxury private car feel. Now, people take Lyfts and Ubers interchangeably, and drivers make money driving for both companies. No peer-to-peer ride company owns its supply or demand, either.

It’s great for those of us who participate on these platforms, because it means the company’s value is concentrated in the people’s, not the corporation’s, hands. If any one of these peer-to-peer companies were to shut down today, we’d just use the next-available platform.

But the corporate entity itself does provide value. They’ve figured out the logistics of the operation, for one. They successfully herded us all onto one platform. They bear the risk of the business. And there are employees, investors, and friends and family who have taken a chance on them.

It seems that successful pure peer-to-peer models have to be monopolies in their respective markets, or else they will be indistinguishable from their competitors at scale. Rather than corporations, perhaps these peer-to-peer models are conduits, not dissimilar to governments, overseeing a certain class of human behavior.

This differs from traditional wisdom about perfectly competitive markets, which suggests that many players and low barriers to entry are a good thing. To compete with a peer-to-peer monopoly, new companies entering the market would need to dismantle it piece-by-piece rather than compete with it head on. (See also: Craigslist killers.)

It seems that peer-to-peer models are here to stay, but we haven’t yet figured out what they will look like at scale. Will they be public corporations, like eBay? Should they be small and profitable private companies, like Craigslist? Or thriving nonprofits, like the Burning Man Project?

(A relevant postscript: if it is true that peer-to-peer models are different than anything we’ve seen at scale, what does that say about the rewards we should expect as investors, and how we finance them moving forward? As Christie George of New Media Ventures puts it, “While the sharing economy is driving forward innovative business models around shared assets, the way we finance these companies is still relatively traditional.”)

*Skillshare and Lyft are Collaborative Fund portfolio companies.