What does Web3 mean for SaaS companies? (Part 1)
An interview with Colin at Protocol Labs
What does Web3 mean for SaaS companies? That was the question I sought to answer in my conversation with my long-time friend Colin Evran, who runs Ecosystem Growth and Operations for Protocol Labs. Protocol Labs has been a primary contributor to the Filecoin Project (the peer to peer or blockchain version of Amazon Web Services) and IPFS (the peer to peer version of the HTTP protocol).
Back in 2017, I went down the rabbit hole of crypto. I was a total geek. At the dinner table, I'd scribble diagrams on napkins about the mechanics of mining. I went to blockchain meet-ups. And I very briefly imagined building a blockchain-based customer success platform while at Gainsight. It was way too early for that – but, the question I explored with Colin was, how early?
Innovations in the crypto world have largely focused on the infrastructure layer. Colin's company has been impactful there. But the application layer is also taking off in the consumer world, and some day the B2B application layer is bound to take off. Applications built on blockchain can afford many benefits: they can increase trust, security, transparency, and the traceability of data shared across a business network – and in some cases deliver cost savings with new efficiencies. When will software purchasing start to shift from SaaS to blockchain, and what will that shift mean for people who work in SaaS?
My preliminary answers: sooner than you think (years not decades), and the new world looks dramatically different from the SaaS world.
I should disclose that I do own some Filecoin, and this post is not intended in any way to promote the token. Please do not use this article as a basis for your personal investment decisions. The purpose of this post is to share with you my learnings about what Web3 means for SaaS companies. And I learned a lot!
I condensed my conversation with Colin into an eight-page Google doc and divided it into two posts. The first is about Filecoin's potential value proposition as a Web3 product versus the Web2 incumbents, i.e. the cloud storage products that are the foundation of the SaaS world. The second post is about how Protocol Labs is organized, as a Web3 company. This conversation is long, but in my opinion, mind-blowing.
I understand that now that you've been at Protocol Labs for a while, you're known as Protocolin. I'm going to call you that from now on.
For us laypeople, what are protocols, and how are you a lab for them?
Here's how I think about it. Back in the 90s, there was Web 1.0, which consisted of static websites that gave you "read" capability. Then in the 2000s through today, you were able to both "read" and "write" through Web2 -- interactive websites like Facebook and others. We're building Web3, which involves "read," "write," and "trust."
Right now you have to trust Facebook to not use your data for bad things, or a government to not censor access to data, or a bank to protect your money. In Web3, you have thousands of computers around the world who are verifying your data, your money, and your agreements (via smart contracts). So you have total trust and verifiability.
Going back to protocols: they're a way of fetching data. We built what we call IPFS as a replacement to HTTP, which is what Web 1.0 and 2.0 were built on. Folks have built all kinds of applications that fetch data using HTTP. But you could rebuild all of those B2C and B2B applications -- like Google or Salesforce or Uber -- on IPFS, in order to ensure total trust.
What does IPFS stand for?
InterPlanetary File System.
I just want to make sure I've got this right…InterPlanetary File System? I plan to vacation on Mars so this may come in handy. Where did the name come from?
I know it sounds futurist, but it’s named that way because IPFS is architected so it can actually work just as well in Space as it can on Earth, whereas existing protocols like HTTP won’t have those properties. IPFS searches for a piece of content regardless of location, whereas HTTP is architected to find a piece of content in a specific location - which constrains its reach.
There are actually some plans to launch IPFS nodes into Space in 2022 - someone created https://areweinterplanetaryyet.org/ to track progress against that goal!
What are some examples that show why I can trust IPFS more than HTTP?
Let's say I want to watch a Netflix video, and separately you also want to watch it. We'll enter in an HTTP address by clicking around on Netflix, and each of us will fetch the same, single copy from a particular datacenter. With HTTP, the copy is fetched from a specific location (for example, a datacenter in Virginia), so if a government or Netflix itself wanted to censor that video, they could do this by blocking off access to that single IP address. China does this all the time. Or if there was a natural disaster and the single copy disappears, that's it – you can't access it anymore.
IPFS says, forget about this single location thing. We can create numerous copies of the video all around the world. All of the copies have the same unique cryptographic hash that can't be replicated unless they have the exact same content. So if one copy is destroyed, I can still access it somewhere else. I can even access it offline through a peer-to-peer connection.
Accessing files using IPFS is actually a lot more efficient. With HTTP, an entire apartment building is downloading the same video from the same datacenter in Virginia at the same. With IPFS, one unit can download the video and redistribute it to other units locally – acting as a decentralized CDN [content delivery network].
What is Filecoin and why does it matter?
Filecoin is the top-layer incentive for storing data on IPFS. Think of Filecoin like the “Airbnb of cloud storage” – where instead of storing data with a handful of corporations, you’re now able to pick from thousands of “hosts” or data centers, each with their own unique characteristics. The current default cloud storage options are AWS, Google Cloud, or Microsoft Azure – users don’t have a lot of choice, and they have pricing departments that decide how much to charge you. They compete with each other, but there's a small number of them, which means you're going to pay more -- in the same way you'd typically pay more for a hotel room than a similar Airbnb. The Filecoin ecosystem is decentralized, so a whole bunch of smaller players can deliver the same benefits as AWS in aggregate, but also compete with each other in terms of retrieval speeds, pricing tiers, geographies, and other attributes. So prices are lower and users have much more choice on who stores their data.
A second benefit of decentralization is that the Filecoin ecosystem gives users full control over their data. When someone is storing your data, they are limited in what they can do with it – eliminating things like selling data for ads without explicit permission.
A third benefit is that there are zettabytes of storage capacity around the world that could be used, and we're not currently tapping it. Just like there was a ton of latent capacity for vacation stays in people's homes, and Airbnb unlocked that capacity.
Filecoin launched in October 2020 and in just over a year, more than 3,700 datacenters are plugged into the network. In total they offer over 14 Exabytes of storage, which is around where Google Cloud was in 2018, or about 1% of total cloud storage in the world today. That's 2,000 Netflix libraries, or almost 4 billion 1080p movies. Storage on Filecoin is also currently less than 1% the cost of AWS.
Who's storing their data here?
The demand side is growing quickly as well. In the early days of iOS, there were a ton of iPhones sold around the world, and Apple figured out how to encourage a whole community of developers to build applications for the App Store. Some of those applications eventually became multi-billion-dollar businesses. We expect the same result from developers storing data on Filecoin.
There are already hundreds of developers across different industries, across B2B and B2C. Most of them are Web3 businesses -- for example:
31 million NFTs are stored on Filecoin across many different blockchains (up from 15M on January 1st)
Web3 Metaverse companies like the MoNa are using Filecoin for storing and retrieving assets
The Brave browser, which is basically a Web3 alternative to Google Chrome or Safari is using IPFS (and soon Filecoin)
Microsoft ION, which is Microsoft's decentralized identity group (think single sign on in the Web2 world) is using IPFS and Netflix has been experimenting with IPFS for faster container image distribution.
The Shoah Foundation, which has a large collection of genocide testimony from survivors, is using Filecoin to store 55,000 videos and 10 petabytes of data. There are a lot of fake genocide videos, so Shoah needs a way of authenticating them. Because each video has a unique cryptographic hash, if you change one pixel of the video, the cryptographic hash will change. So once something's been authenticated, you can trust it. At the front end, the foundation has a mobile app that stores metadata on who recorded the video, when, and where. Imagine what this technology could do to eliminate fake news. The same technology has been used for preserving the authenticity of the January 6th Capitol Riot Images.
If I'm a Web2 SaaS company, why would I store data on Filecoin?
Yes, you'd use Filecoin for a few different reasons. First, Filecoin is 99.5% cheaper than AWS right now, because of the competitive dynamics I mentioned earlier. So for companies with extremely large data sets -- like genomics data, or climate change pictures (often taken from space) -- Filecoin is a great solution. For example, Columbia University (Max Planck Institute) is making Petabytes of climate data available on Filecoin to climate researchers, and NYC is putting air quality, city demographics and public hearing data on Filecoin as well. Even for an average SaaS company, there are major cost savings.
Second, if you're a datacenter storing user data on Filecoin, you're actually earning Filecoin tokens, in the same way that Bitcoin miners earn Bitcoin for proving the consensus mechanism on the blockchain. We call it a "block reward," which is an extra cost saving -- basically a subsidy, which is really high today although will decrease exponentially in the future as more and more Filecoin are created. So 30 years from now, we would expect 99% of a SaaS company's fees to come in through client fees and 1% from block reward. But right now it's probably like 99% block reward, 1% client fees.
A third benefit is that SaaS companies currently have vendor lock-in with AWS or equivalent. One you tie yourself to one cloud storage provider, you can't easily move your data to another. But you can do that easily on the Filecoin network.
Fourth, SaaS companies get the benefit of verifiability. If a SaaS company's users want to control their data, rather than trusting the SaaS company's promises about how they're using and not using that data, those users may pressure the SaaS company to store their data on Filecoin. On Filecoin, no one can access your data without your private keys, so as a user, you get full control over your own data.
Offline access is a fifth benefit. There's a company I'm a big fan of called Huddle that does video conferencing like Zoom but using IPFS and Filecoin. They are focused on education in rural India, where internet connectivity is really poor, and it's expensive to store data. They're using IPFS to move data from a video call within local area networks, rather than a connection to the internet backbone, and store data cheaply on the Filecoin network.
I know Filecoin is open source. Can you describe your ecosystem?
If we went bankrupt or disappeared or decided not to do this anymore, all the code is open source. There are actually 8,000 other developers apart from our team that have contributed to this code, and anyone can fork the code and take it in their own direction as well. We're one of many who are contributing to the ecosystem. We put together the initial implementation but we're a small part of it at this point. It's kind of like Rust or Linux where there are different working groups of other folks that help shepherd the protocol. Same thing with Bitcoin - there's no central company for it. It's owned by the community. We don't even know who created it!
So how do you run a company in service of an ecosystem?
…To be continued.
Stay tuned for part 2 of this conversation, coming later this week. We'll cover: How does Protocol Labs operate differently from a Web2 company?
If you'd like to hear from more tech leaders like Colin a few times a month, sign up for my newsletter here: