The Virtual Real Estate Myth (revisited)
(In November 2021 I published a paywalled article warning people not to invest in metaverse “properties”. Now that prices for these digital assets have crashed1, I’m making a version of text, along with some new thoughts, available to everyone. I’m doing this in order to gloat (who doesn’t love to gloat?), but also to give free subscribers a taste of what you get for going premium. If you saw the original version of this, you should still read on as I’ve clarified some examples and added a section on how to approach new asset classes generally. I don’t want to give too much away, but there will be talk of webrings, Weblo, CryptoKitties, Hot Topic, and Dubai’s slowly sinking islands).
There is no such thing as virtual real estate
Let me start this analysis by talking about the wrong way to view metaverse “real estate”. The reason real-world real estate works how it does is because of three basic principles:
1. Scarcity.
2. Proximity.
3. Inherent value
The second one is the one you usually hear about, as in “location location location”. The first one is implied for any real world asset, so no one says it. Sometimes you hear it stated explicitly for a subset of real estate, like the mostly true statement that “they're not making any more waterfront” property (and FWIW, all three of my most recent house purchases were on waterfront).
Principle 3 has to do with things like oil deposits. The closest digital analogy would be with owning tokens on a proof-of-stake blockchain, not virtual real estate. I won't be discussing it. It also has to do with the fact that you can live in (residential) real estate, and everyone has to live somewhere. If you see value as not entirely subjective or subject to changing whims, then you have to recognize a spectrum, from inherently valuable assets like precious metals, to items with completely subjective value, like CryptoKitties. Real estate, depending on the specifics, can be close to the inherent value end of the spectrum.
For the moment set aside inherent value, and just put principles 2 and 1 together. From those, you see that real estate buying decisions are constrained by the desire everyone has to be close to a desirable location (or directly in one), and by the fact that space close to those locations is always limited.
Now let's look at a VR world.
It could be laid out just like our world, with virtual streets of rows of houses, and virtual malls with rows of stores. And some metaverses are laid out like this. For now. But let’s take a step back and look at this from a competitive marketplace point of view.
If you're fat IRL, will your avatar struggle to go up stairs?
If I can choose between shopping in a metaverse where I can instantly "apparate" (Harry Potter style, but perhaps without the squeezing and puking) to any other store, or a metaverse where I have to slowly walk from TJ Max to Hot Topic, which would I choose? Keep in mind that the one with apparition is also easier to program, since you don't have to replicate the experience and limitations of walking through mall corridors.
Maybe some people will prefer limited choice and slow navigation, but I think it's telling that all our TV remotes have number pads, not just up and down buttons to honor the past experience of turning the dial through intermediate channels.
What will make sense are recommendation engines and targeted ads. If I'm in a metaverse that's a managed, walled-garden, experience, and I've been shopping for shoes (real or digital) at Hugo Boss, an ad might pop-up inviting me to pop-over to the Salvatore Ferragamo store and level up my kicks game. If you really want to make money, figure out how to redirect people searching for "shoes" to the brands they really want; I did this kind of "Adsense arbitrage" for a while on the regular internet and it was primo good. (If you want to try a related strategy in a metaverse which you understand well, send me an email and we'll combine brains and clean up.)
But I have neighbors in my digital world!
One objection to my objection is that, as mentioned, we already have virtual neighborhoods with virtual neighbors. People are buying plots and selling plots based on what's “next” to them, right now!
To which I say: remember webrings? And for those who don't remember webrings, they were part of the solution to the problem of discoverability in the early days of the internet. You'd join together with other webmasters and run some code on your website that linked to the “next” website in the ring. Webrings still exist, but if you still use them you are very much an outlier. Everyone else uses search engines, bookmarks, or if they want fun and quirky sites they use any number of aggregators, like pointlesssites.com (not linked because you really shouldn’t).
In the real world, proximity (location) is mindshare. I notice the Pretzel Hut because it’s in the mall where I’m shopping already. Pretzel Hut does’t have to advertise because it’s already paid for proximity (including proximity to their own carefully crafted and wafted smell); they get noticed because shoppers see and smell them.
None of this maps cleanly onto the virtual space!
Not convinced? Can you come up with any mature, multi-vendor digital space where locations and proximity map the same way they do in the real world? Can the iOS app you built have a permeant home next to another specific one on the App Store? If you are playing Skrym in Steam, and you wander to the easternmost edge of that virtual world, do you catch a glimpse of Halo?
Investors have a long history of going bankrupt when they import the model from one context into another and think the same dynamics will hold. I've made that mistake myself and lost 50 grand and two years of effort.
In the short run, you might find some lovely arbitrage opportunities in flipping fake properties. In the long run, having the best "location" in a webring is worth exactly what you'd expect it to be today.
Prestige, flexing, and NPCs
One more note about digital goods and prestige.
Prestige is always scarce, but virtual prestige can be unlimited. Remember that scene from Minority Report where some guy is plugged into a memory machine and everyone is telling him how admirable he is? THIS is what people will pay for. And for virtual sex, of course, the more realistic the better. Neither of these is inherently scarce if provided by algorithm. People will pay more for more "premium" virtual experiences, but once coded, there's no limit to how many of those experiences can be sold. And if a vendor artificially limits the number of "you are god-king-rock-star-sex-idol" fantasy experiences they sell, someone else will code a similar experience, or reverse engineer it, or pirate the code, or whatever. Unless people give real-world praise to people who purchase artificially scarce virtual experiences from a specific vendor, there's no reason to overpay for the exact same thing.
To be as clear as possible, we have no long-term example of software code being sold in a rate limited way. And before you yell “NFTs”, keep in mind that 1. These map to collectables more than code or 2. Many of these are usage tokens, which have value because they are like Chuck-E-Cheese arcade coins. You need to buy them to participate in some activity within a walled garden. Ether (the “gas” that powers the Ethereum blockchain) has value because it has to be scarce, and valuable, to work as a rate-limiter. My use of Photoshop isn’t diminished by it’s ubiquity, and in fact is only enhance because of network effects from other users.
What will have value
So metaverse “real estate” is out, but what's in? Besides the potential visitor arb opportunities I mentioned, I would focus on three things if you still want to invest in a meta world:
1. Tools. These could be headset manufacturers, or software frameworks for building things in metaverses. You know, the whole sell shovels to the gold rush prospectors strategy.
2. Discoverability. Metaverses have the same problem that the internet has: infinite stuff, but it’s hard to find what you want, and even harder when you have multiple metas. If you think about the main driver of every massive content sharing site to come of age post-Facebook, every single one depends on a kick-ass algorithm for presenting an endless stream of semi-passive entertainment. (How could Tik-Tok map to a metaverse?)
3. Marketing and advertising firms that get the space.
4. Coins/tokens if you have the stomach for a very rocky ride in terms of prices and maybe even liquidity.
Understanding new asset classes
Like many a web developer in the late 90’s and early aughts, I decided I needed to learn XML. It took me a surprising long time to wrap my head around this new web standard. I kept trying to understand it in terms of things I already knew. Was it like a text form of a database? No. Is it a spreadsheet in markup form? Most definitely not. Was it HTML for data? Not quite. To finally understand XML I had to let go of the comparisons I had in my head. XML is it’s own thing. At the highest level of abstraction, it’s a way to organize information according to certain formatting constraints, which is also true of the things I was comparing it to, but to understand what makes XML XML, you have to stop trying to view it as a modified version of a spreadsheet.
“Virtual real estate” isn’t like “virtual” real estate. It’s not even remotely the same. To the extent that any of the three principles of actual real estate apply, that’s because the metaverse designers have tried to artificially introduce them into the design of their metaverse. But this is like designing an operating system that only allows you to move files by drag and drop, because that’s what we’re limited to with real files on a physical desktop. Even before the current round of metaverses, there have been multiple failed attempts to sell assets that mapped to real places, from the embarrassingly named Weblo, to Dubai’s slowing eroding islands in the shape of the world, unimaginatively named The World2. The only successful metaverses are simulations, usually in a game. These sell the experience, not the houses.
At the risk of overemphasizing this point, you simply cannot understand an asset class based on what it is supposed to map to, or based on the high-level metaphors used to describe it. Is bitcoin mining like gold mining? Only in the most abstract sense that machines and energy are used to gain control over an asset of some kind. But trying to understand blockchain or hashing algorithms or distributed consensus by comparing these things to pulling gold out of the ground, that’s only going to cause confusion. Even in terms of economics, digital tokens aren’t digital gold! The only thing they have in common is list risk of hyperinflation than fiat currency, but even here the dynamics are way different. Technology that could massively increase gold production is very different from a human decision to alter the bitcoin algorithm in favor of more new coins.
From a recent The Wall Street Journal article: “Meanwhile, the price for virtual real estate in some online worlds, where users can hang out as avatars, has cratered. The median sale price for land in Decentraland has declined almost 90% from a year ago, according to WeMeta, a site that tracks land sales in the metaverse.”
Math nerd fact: Any map of the real world, when laid out, must have a point on the map that is exactly on top of that same point in the real world.