Pocket Network vs. $17 billion of startups
April 21st, 2022
A confusing part about web3 is whether or not any centralized entity is allowed to control anything.
Purists would say no, I suppose: it’s decentralized all the way down or bust. Chris Dixon would say yes: centralized entities can exist and can offer better experiences (it’s just that decentralized data keeps them from doing monopolistic things).
In that sense, the fight that Pocket Network brings to Infura and Alchemy in the realm of blockchain access represents a test. Can a decentralized entity actually provide better service than a centralized one?
Accessing the blockchain
Here's an oversimplified way to understand how a lot of software that you use actually runs in the background:
A database holds a bunch of data like names of people, their favorite instrument and how many bananas they've eaten
A server helps you pull the right data (“the top 10 banana eaters” from the database and spit it out in the right format
A front-end shows that data in the pretty way that you like (a sorted bar chart of banana eaters)
So when you visit myfunapp.com/dashboard, this happens:
Server gets a request to pull data
Server sends a request to the database
Database returns the data to the server
Server spits out some formatted info
Front-end shows it in the dashboard
If you’re crafty enough with your tooling, you can figure out that visiting api.myfunapp.com/users shows you the data from step 4, before the front-end makes it pretty.
All of that is built and owned by the company, hosted by a cloud provider (Amazon, Google, Microsoft).
In particular, the company made sure to provide an access point (the link that shows the data from step 4) that lets its dashboard get data. And if the company’s mission is to make it easier for other companies to build products, it’ll make that access point really accessible to everyone, not just its dashboard, and say, “Hey, here’s where to go if you need to access this data.”
An oversimplified way to define a blockchain is "a database shared across many nodes". Instead of the database living in one place run by Amazon/Google/Microsoft, the database is copied onto many different computers owned by many different people.
The first important question of the day: who do you ask if you want to get data from the blockchain? Someone needs to write the server functionality that allows you, a developer, to get data and do stuff on the blockchain via code. A company did that in the previous example - there’s no “company” here.
If you wanted access to the data, you’d have two choices:
Run a node yourself and build in the functionality that reads and spits out the data
Convince an existing node to do that, presumably in exchange for some money.
We know the former doesn't really work at scale anymore - that's why Amazon and Google and Microsoft are making boatloads on cloud services. And the second, well, that's exactly what happened.
Infura and Alchemy
Today, if you're a developer wanting to build apps on top of the blockchain, you go to either Infura (owned by ConsenSys, valued at $7 billion) or Alchemy (valued at $10 billion), you ask them for data, and you pay them some amount of money per month.
What they do: Infura and Alchemy each run their own nodes on various chains, field those requests and do things on the blockchain in exchange. They might also build some extra functionality on top to improve the user experience, but fundamentally this is the business they’re in.
Pretty much every single application uses them. They’re easy to use! They provide a good experience and most developers are familiar with the model.
You’re probably using Infura: It happens to be that MetaMask, the most popular wallet used by 20+ million people every month, is also owned by ConsenSys. When you want to do something on Ethereum, MetaMask uses Infura under the hood.
MetaMask simply forwards whatever it is that you want to do along to Infura to handle it for you. Here's a screenshot of the network settings that MetaMask preconfigures for you:
The second important question of the day: if the blockchain is decentralized, should access to the blockchain also be decentralized? Is it not at least ironic and at worst problematic that doing anything on the blockchain currently runs through two companies?
If your answer to either question is "yes", then Pocket Network will resonate with you.
Pocket Network, a decentralized competitor
At its core, Pocket Network matches up two parties: (1) applications that want to interact with the blockchain and (2) nodes that want to provide the service.
Less cost: Developers pay a one-time fee to get a certain amount of requests per day. That means it’s often cheaper to use Pocket in the long-run since Infura/Alchemy both charge recurring fees per month.
When the application wants to send a request, Pocket will then pick from its network of thousands of nodes to pair with the application. The node will service the request and get paid in POKT.
Less downtime: By decentralizing access, Pocket also hopes to improve reliability. Instead of relying on nodes from the same provider (that also probably use the same underlying infrastructure), Pocket allows developers to plug into a variety of nodes to minimize any downtime.
Today, Pocket Network has over 40,000 nodes online across 13 different networks and supplies close to 800 million requests per day.
Pocket’s spotlight moment has been helping Harmony Network scale as it reached mainstream awareness.
Harmony Network is an Ethereum competitor that, for the most part, was one of the many "ghostchains", a chain with not much going on in the ecosystem. The blockchain existed, but there wasn't really much to do on it.
That started to change in Spring/Summer of 2021, when a game called DeFi Kingdoms launched on Harmony. Crypto Twitter loved it - it reeked of potential. And so by Fall of 2021, DeFi Kingdoms started driving a lot of traffic to Harmony.
The spike in popularity is a sign of how quickly things can change in crypto and how infrastructure can be caught off guard in those moments.
Harmony hadn't ever been tested with this type of volume before, so soon, the game would start breaking. You wouldn’t be able to login, the game wouldn’t show you how many game tokens you had, etc. DeFi Kingdom's popularity was overwhelming the nodes handling the users' requests.
Harmony turned to Pocket Network to help.
Backed by its community amplifying the message, Harmony began instructing its users to configure their wallets to send transactions not to Harmony's default settings but to Pocket's Harmony nodes.
It also redirected the requests it was receiving via those default settings to Pocket Network behind the scenes, which helped spread the load to more nodes. Pocket Network handled nearly 8 billion requests for Harmony in January 2022.
In or out of Pocket?
So, here’s my take on Pocket:
I want something like Pocket to exist. It’s just weird to me if there are only two companies that do well and no protocols.
For now, Pocket’s success is mostly riding from cultural preferences. That is, developers build on Pocket because they believe in decentralization.
Developer audiences are still very small. The cost and downtime angles could prove to be compelling given just one or two publicized instances of Pocket outperforming. One major project saving massively on spend or staying alive when everyone else goes down could be all it takes to shift developer attention to Pocket as a default.
From a distribution angle, it has to get very sharp on winning developers, which it’s not right now.
From a product angle, it’s also not clear if Pocket’s scope is actually enough to keep developers in the long-term. There might be other services that centralized entities like Infura/Alchemy can provide, that developers need, and that Pocket just can’t handle.
There are a lot of questions facing Pocket moving forward. Overall, I’m thankful that Pocket exists and is presenting the challenge to the space!
What’d you think of this week’s update?