Protocols are a geeky topic. It's way more interesting to talk about what you can do with a protocol. If you're Fred Wilson, apparently it's more interesting to talk about how you can monetize the use of protocols. This post is going to be a response to his recent post, The Golden Age of Open Protocols.
As the central topic of Fred's post is open protocols I think it would be prudent to define what an open protocol is. While there there is no single, true, definition, I think that there are a few core aspects that are nearly tautological.
An open protocol must be two things, standardized and open. "Standardized" is easy, whether you have a strictly defined standard that spells out every minute detail or a webpage the the gives the bare minimum details of an API, there must be something that tells others how to use that protocol, and your services or interfaces must adhere to those standards.
"Open" is a much harder term to pin down exactly what it means. To some, it can just mean making the document that lays out the standard available in some form or another, to others, it can mean a very specific set of processes that have to be gone through to arrive at that standard and a specific set of requirements for how that standard is made available and what it required to actually implement it.
At an absolute minimum, I think it's safe to say that, to really be an open protocol the standard specification must me made freely available, not requiring any payment for access, payment of royalties to implement, or paid licensing of patents held by the creators.
There are some other factors that really get at what the spirit of open protocols are all about that are a bit more nuanced, we'll get to that in just a second. Getting back to the original post, the key piece of where I'd like to start this discussion is the following quote:
One of the problems we have had in tech is that there aren’t large monetary incentives to create and sustain open protocols. If they are open they cannot be easily monetized by traditional means.
This is really the thesis of his post and is definitely the point where we most disagree. Protocols themselves should not and likely cannot be monetized. First I'll start with why they should not. In the paragraph before the above quote, Fred says:
Open protocols are at the heart of many of the most important systems that we have. The Internet works because of TCP/IP. The web works because of HTTP. Email works because of SMTP. These are open systems that developers can build applications on top of.
This is all very true. Without these protocols the world wouldn't look like what it does today. However I think the rest of the post misses out on just what makes open protocols like HTTP or SMTP as useful as they are.
These protocols wouldn't have seen such a wide adoption if they used a system like what is suggested. With HTTP, I can read the standard, write my own server from scratch, and make it available on my local network. I can then write my own HTTP client program to access that local HTTP server. I can do both of those things without interacting with any outside services. I can then make that server available to the greater internet and, assuming I implemented it correctly, anyone else can then access my server with any other program that implements HTTP as a client. This can all without paying anyone¹ or asking anyone for permission to do so. This is what causes the explosive growth that open protocols create.
I also mentioned that I think that protocols likely cannot be monetized. The problem with trying to monetize a protocol directly is that any form of required payment to use the protocol (as opposed to required payment to use a service) is that the payment is nothing but overhead. A totally equivalent alternate instance of the protocol could be started (the equivalent of an alt-coin) with lower costs to use with zero degradation of the application functionality.
Beyond that a completely different protocol could be created that does not depend on a blockchain to provide exactly the same functionality to the end user. This means that the addition of the blockchain crypto token system is only adding overhead and because of the way a blockchain has to work it is quite a significant overhead indeed.
The reality is that the addition of payments and value to the protocol itself has much more benefit for those trying to make money off of it than it does for those trying to use it. For those that do just want to use the protocol or the applications that would be built on top of it, it is nothing more than very complicated overhead that makes systems less reliable and much harder to reason about. Without any significant benefit to the users of the protocol, there are no users, and without any users, there's no way to make money off of it.
The original post seems to be saying that we're on the cusp of a golden age for open protocols, but the truth is we're already in that golden age and have been for some time. The whole of the internet is built on layer after layer of open protocols and it doesn't seem to have been suffering from not being able to directly monetize each protocol's use. Really it seems quite the opposite, that the availability of truly open and free protocols have caused the internet to become the ubiquitous foundation of modern society that it is today.
1: I'm obviously omitting any need for paying for hardware or internet service as this is going to be required no matter what.