Where the Real Network Effects Are

The Question
The most common single sentence dismissal of Polkadot I see on Crypto Twitter is “It has no users”. This take comes in two main variants:

  • The “things won’t change” argument: not many users now = not many users ever
  • The network effects argument: not many users now = substantial hurdle to adoption where other networks have advocates, liquidity, and TVL based network security.

This second argument begs a few questions. Which network effects matter? Which things do we think of as network effects that aren’t at all? Which network effects are weak, and can be overcome by a marginally better tech offering? Which network effects are mission critical to cultivate, and therefore must be optimized for at an early stage?

A Starting Place
I’ll leave this as more of a discussion than a monolithic opinion post, but first here’s a list of some perceived network effects and where I think they rank in terms of importance to Polkadot’s success. I’m very open to arguments in favor of different importance weightings or disagreement about whether one of the below is a network effect :slight_smile:


  • Developer activity in ecosystem


  • Developer activity in ecosystem languages
  • Number and reach of active advocates
  • User base/liquidity in hosted NFT auction houses
  • TVL in staking protocols for POS or equipment/electricity cost for POW


  • User base/liquidity in hosted fungible token exchanges
  • Number of users with active wallets (This may become important as peer to peer transactions take off as a use case
  • Unique users in social graph

Not a network effect

  • TVL and user count for hosted dapps where each user only interacts with a central authority or isolated contract (EX: dapp that scrapes chain for requested data)

I think one of the common traps here is to compare the Polkadot relay chain to other existing layer 1 networks.

Polkadot is more of a layer 0 mulit-chain protocol, of which a blockchain is needed to provide value to the parachains built on it. The Relay Chain’s purpose is very meta, and mostly does not involve or provide value to “users”, so if we try to compare it to the kinds of metrics which we look for in layer 1 blockchains, certainly it will fall flat.

Here is a definition that @gavofyork wrote for the academy about the “blockchain layers”:

Polkadot is a Layer 0 Blockchain

Layer Gav’s Description
Layer 3 An application or platform which is hosted by an L2.
Layer 2 An application or platform which is hosted by an L1.
Layer 1 A basic blockchain which administers a native token and hosts one or more applications.
Layer 0 A secure consensus-provider. A pure L0 has no ability to host applications. An impure L0 may be able to host limited fixed-function applications (e.g. token hosting), but is primarily used to host other blockchains.

So really we need to look at the parachains in aggregate and their activity when trying to make comparisons to basically any other blockchain. We must also take into account the lifetime of parachains compared to other ecosystems. Yes, Polkadot has existed for 2 years now, but the first Parachains were onboarded only on December 18th, and most don’t actually become fully usable until a while after that.

Anyway, these are just things to consider when comparisons are made.

To your question:

I think our strategy has been the following:

  1. Build a truly unstoppable and scalable application platform for Web3 developers.
  2. Build a framework for easily developing modular and extensible application-specific blockchains.
  3. Educating and nurturing teams to build successful and high quality projects on this framework.
  4. Empower and support teams to build out their communities and users, and bring them into the overall Polkadot Ecosystem.
    • ??

As you can see, I think we are on phase 3 of this mission, and you can see that there has been A LOT of output trying to specifically support teams and individual developers building in our space. Just in the last 6 months or so, academy, stackexchange, forum, and improvements to the developer hub all came to be.

I do not believe we are there yet in building enough successful teams, with enough quality and knowledgeable engineers to provide great applications to end users. I think if you asked any parachain team, they would easily ask for double the amount of resources, and Parity might answer the same about that too.

This is why, the only real metrics that I focus on at this moment are the number of developers building in the Polkadot ecosystem and GitHub activity. If we can be successful and continuing to build this out, I think naturally we will build up the end users that you state is lacking.

That all being said, I struggled to think of a concrete set of ideas or plans around how we could best accomplish phase 4. I think it starts with identifying and supporting specific kinds of applications which we think have a good intersection of being needed by the world and being particularly well supported by the Polkadot protocol.

What does that look like to the parachain teams which exist today?


Very good points! Your post really speaks to how Polkadot is building out network effects starting with the most important one, developer activity. There’s a virtuous feedback loop here where more developers result in better tutorials, StackExchange posts, development tool quality, etc.

With Polkadot I think it’s also appropriate to divide the concept of a developer network into separate layers per height in the tech stack.

  • Substrate/Node developers
  • Frame/Parachain developers
  • Smart contract/dapp developers
  • Front end developers

Parachains seem to be reaching maturity as test beds such that we could see a tipping point in dapp developer and front end developer interest. I for one am very excited to see the ease of product development that could result from a strong network effect in dapp engineers!

To answer Shawn’s question from the Parachain builder’s perspective, let me add my short comment here.

The strategies written above (1-4) are great. But what we are missing right now in terms of the network effect is to make a Parachain’s use case that leverage other Parachains’ features. For example, Astar can use OAK’s automation to make Astar’s dApp staking auto-compounding. I think Parity’s marketing team needs to highlight the power of XCMP more.

For me, I think parachain teams including us are satisfied with just opening HRMP channels. As a result, what we are making right now is just to bring Parachain A’s token to Parachain B and vice versa. Unfortunately, that’s it! To make a strong network effect, we need to integrate other Parachains’ features into our own Parachain just like an Ethereum smart contract integrates other smart contracts.

As a founder of Astar, we will make more use cases that focus on XCMP in Q4 and Q1 2023 and we want to make showcase projects together with Parity and ask for guidance somehow since Parity knows XCM best in the ecosystem.


Network effects and adoption comes when you have a proposition that appeals to many end users and their friends.

Parachain teams probably should adopt a start-up mentality towards user growth even if the eventual aim is to have their work long-lived and fully decentralised. That should help inform how things are built and tested even to the point of asking “how do we get one user to engage their friends?” - the classic social proof problem.

I’m not even sure that Polkadot as a layer 0 needs to be part of that discussion because a well known or popular Parachain will have a symbiotic relationship with Polkadot anyway regardless of how they promote their own product.

1 Like

I agree with the “Critical” point from @brad.olson.587.

The entry bar to polkadot ecosystem is a little bit high. We really want to see some pet projects to appear on the network, but substrate learning curve with all its beauty of modularity is still steep in comparison with smart contracts

And this is why ink! exists in the first place in my opinion. We want to lower entry bar as much as possible and provide a smooth transition from the smart contract development into the substrate.

Adding tutorials and workshops explaining how to transform the smart contract into the substrate pallet might be one of the options.

However, as a student, I notice that polkadot brand awareness is still lower than of alternative protocols. Newcomers in web3 still consider learning either solidity for Polygon, Avalanche or rust for NEAR (I don’t really have metrics to back up my points, but this is the personal impression I get from talking to different people).

What we really want to do is to build a developer momentum around polkadot and its tooling. We have beautiful tooling, some people just never heard of it, and we need to fix this.


Playing devil’s advocate, what if that alone was the killer feature of the technology?

Afterall, what you’re aiming for is uniquely possible within the domain of a general purpose smart contract chain, but is not optimised for Polkadot’s sharded design?

agree with most here, adding possibly a different angle.
Historically, individual cryptos got network effects at some point, and if they survived, then they stayed at very top in almost cemented way. Over time, those cryptos re-adjusted narrative or even product to stay at the very top somehow and keep attention flowing (btc forks, ETH pos merge etc.). You probably can’t win the onboarding game, since their network effects are larger.

Let’s look at average journey of a user:

  1. buy bitcoin on coinbase (0-1 years)
  2. buy eth and go dApps (1-2 years)
  3. learn more about advanced stuff, and move further (cosmos? polkadot?)

Average dev goes through similar journey, where they start with some popular L1 smart contract devving - ETH, Solana maybe.

The key point to adoption is right targeting of audience, and it seems it’s not about new comers to crypto. Polkadot should target crypto natives that are in the space 1-2 years and are already familiar with concepts, so they can see value in USPs of Polkadot.

e.g. - Solidity devs, Solana devs, Cosmos folks, ETH 2.0 developers that have just begun to think in cross-shard consequences!

In this context, the Polkadot target audience is still very small and is in it’s infancy, but it’s not non-existent. We need to probably start making comparisons and advantages in between ETH2.0, Polkadot, Cosmos, maybe Avalanche and in return you get the builders from those areas. And obviously address the rollups/data availability/celestia paradigm.