Protocols as a service
Can web3 disrupt SaaS?
During the past decade software as a service (SaaS) emerged as the dominant business model for technology startups. The ease with which products could be launched provided a blueprint for entrepreneurs to launch businesses that could readily scale using cloud infrastructure to meet user demand without hefty up-front investment in infrastructure as had once been the case.
The SaaS business model utilises a traditional web2 proprietary platform approach that monetizes via a subscription. This is powered by cloud infrastructure provided by the likes of Amazon and Microsoft, with the SaaS business creating its own silos of customer data.
This is no different to the approach taken by many of the major web2 tech companies such as Facebook, Twitter and Google, except that they operate at a scale that is many multiples larger than most SaaS businesses.
The issue with the SaaS business model is that it rewards centralisation. As SaaS companies grow they become more centralised. This isn't necessarily a bad thing if the competitive landscape remains diverse. However, market segments usually end up being dominated by a handful of key participants.
There is no incentive for these firms to provide interoperability between their platforms, so we end up with walled gardens around the various SaaS platforms.
SaaS is so attractive for businesses as it provides a recurring revenue stream, in many cases with customers paying for services for a whole year in advance. Funds that are in effect an interest-free loan from the customer, which the provider can use to invest in their platform.
It also uses a highly scalable platform model, where once the underlying platform is built out, new customers can be onboarded easily, and the platform can scale — although different phases of scale do provide different challenges for the company.
It's easy to see why most startups focus on becoming SaaS businesses. Hence, we're also seeing web3 companies launching with SaaS models — our own Epirus analytics platform is one such example, where customers pay a subscription for us to run an instance on our behalf.
This is unsurprising given that many of the most successful web3 companies provide platforms that onboard users to web3 in a more traditional web2 manner.
Centralised exchanges such as Binance and Coinbase, whilst not strictly SaaS businesses (as they don't charge a subscription fee) exemplify this approach as to what it looks like at scale. Where you have ever-increasing amounts of centralisation for web3 services, not unlike what happened with web2.
Infrastructure providers, such as Alchemy and Infura, also demonstrate this trend. They simplify access to web3 networks by providing the node infrastructure that is required to run and manage web3 applications.
The issue with these services is that many web3 applications and protocols have come to heavily rely on them instead of running their own node infrastructure. Hence outages to them can cause significant issues to their users when accessing them via websites as the rails their users use are no longer available.
In this respect, the SaaS business model fails web3. If you don't have a service that can provide similar availability guarantees to the underlying blockchain platform they are running you're constrained by the siloed platforms that are commonplace in web2.
To address this SaaS platforms need to exist as decentralised services or protocols. Where multiple organisations can run the core infrastructure of that platform in a location of their choosing.
This approach has parallels with what a number of companies do with open-source software, where they take an open-source code base and offer it up as a component of their proprietary platform, or host it on behalf of customers.
The issue with both approaches is that the creator of the SaaS platform or OSS can struggle to monetise their services if they’re using open-source technology. If you wish to decentralise your SaaS platform you need to ensure there is an incentive for participants to feel compelled to support its decentralisation platform and that you're able to maintain a level of control so that you are not at risk of cannibalising your customer base.
In web3, the most obvious approach is to provide token-based incentives. Where companies providing core services are rewarded with a token for their service to your customer base. This is not unlike how decentralised exchanges such as Uniswap reward those individuals providing liquidity to the pools on their platform.
The bigger challenge is how one can incentivise anyone with a lucrative SaaS or web2 platform to go down this route, and how to ensure sufficient momentum can be created for such services.
Coinbase should certainly be commended for finding avenues to provide decentralised infrastructure, which they are doing through embracing ENS and providing non-custodial wallet infrastructure. This enables them to gradually onboard more of their customers to native web3 applications and platforms.
Web3 infrastructure provider Infura also recently announced they were going to decentralise their platform. Infura and Alchemy are the dominant web3 infrastructure providers, both valued as unicorns, at over $7bn and $10bn respectively (although Infura's valuation is included with Metamask as part of ConsenSys').
In building a decentralised network for infrastructure providers, Infura will need to incentivise providers to join the network. This will likely be with a token to reward participants. The upside is that they will remove the centralisation inherent to their current model providing a native web3 experience for their users.
If Infura does provide its own token for this network, the price volatility will likely put off some participants. However, given the current web3 landscape, it seems more likely than trying to work with stablecoins which do not have the same potential upside as a token.
I highlighted previously projects creating native web3 stacks. Included in this were decentralised platforms for infrastructure services such as data, cloud computing and video streaming. This transition by Infura will position them as another true web3 infrastructure service alongside those.
However, unlike some of the other web3 infrastructure services, Infura has significant adoption from leading web3 applications and protocols. Therefore it could become the first native service running at scale which would be a great milestone for the web3 industry.
It will be important that end users don't end up paying an additional premium for using Infura's decentralised service compared with what they would have before with Infura, or versus a competitor service such as Alchemy. If web3 infrastructure can find ways to be cheaper than their centralized equivalents this will be a huge engine for growth among web3 projects.
Should Infura strike the right balance here, it could become a blueprint for other SaaS platforms to follow with respect to web3. Over time, perhaps this could be the approach that challenges the SaaS businesses of today, with SaaS platforms being replaced by these services which exist as protocols.
This wouldn’t just be good for the web3 industry, but more importantly, the customers and end-users of these services, as they will end up with platforms that are both cheaper, more resilient and provide better privacy of their data than what they have currently, which is worth getting excited about.