If you would like to view the full recording of the live State of Livepeer Q4 call, you can find it on our YouTube channel. You can also read the full Quarterly Report by Mihai Grigore and Stephanie Dunbar here.
Participants:
Mihai Grigore – Messari Sr. Research Analyst
Stephanie Dunbar – Messari Research Analyst
Doug Petkanics – Livepeer CEO & Co-founder
Stephanie Dunbar (0:00): This will be our fourth quarterly call with Messari Livepeer. I’m joined today by Mihai Grigore, my co-analyst, as well as Doug Petkanics from Livepeer. If you guys want to give us a little intro and then we’ll get started with some questions.
Mihai Grigore (0:18): Sure, happy too! Hello everyone, my name is Mihai Grigore. I am a Senior Analyst at Messari and I cover infrastructure as well as base protocols and consumer applications and today we’re very happy to be joined by Doug Petkanics the co-founder of Livepeer. Doug offered to you for a quick intro.
Doug Petkanics (0:40): Sure. Thanks, Mihai. Thanks, Stephanie. Thanks everyone for joining our quarterly analyst call here. I’m Doug Petkanics, founder of Livepeer, where we’re building the world’s open video infrastructure built on Ethereum now migrated in Arbitrum, been alive for six years, so we have a lot, a lot to cover here on the call.
Stephanie Dunbar (1:00): Yes so, this quarter was a big quarter for Livepeer. I believe 42 million minutes of videos transcoded so that was the biggest year to date. And also, this year we got a better look at a little more granular data as to where that usage is coming from which I know was a big ask from the community. Now 90% or more or less of that usage is coming from Livepeer studio so is assisted from the core team’s infrastructure although that is changing over time as well but within that 90% segment, we saw that 36% were from social apps and 40% were from music streaming apps. Now I know we’ve talked a lot about the killer web free social apps that are coming to eat the world in which we expect usage to grow a lot within the Livepeer Community. But I was also really excited and interested to see that using music streaming made up a large percent of usage this quarter so I would love to hear from you what are some of the music streaming apps that you’re aware of that are using Livepeer and how does that fit into the wider web3 social picture.
Doug Petkanics (2:10): Yeah thanks Stephanie! For a little bit of background, you reference the minutes of video transcoded by the Livepeer Network that’s always been sort of the best estimated proxy of demand and usage for Livepeer. Livepeer is a video infrastructure network, so developers use it in order to power the video streaming and video uploads within their applications and as long as the project continues to grow the number of minutes streamed and processed on the network that means more fees are flowing through and more work for the node operators to do and the project is succeeding at becoming a more and more useful video infrastructure. I think you hit on two categories that have been really exciting to us and they’re related music centric streaming web3 social which we think it often as kind of creators creating content and putting it out there to their communities’ followers more so than just kind of one-to-one social networking like maybe in the last era of social. These are two really exciting areas for a few different reasons music I think has high potential because the reason it represents a lot of usage on the network by minutes is because often you can think of things like radio stations and online DJ sets and concerts which typically are very long-running streams sometimes even 24/7 streams that never stop broadcasting in the case of like a digital internet first radio station and so you see apps like The Lot Radio for example that are streaming what’s going on in special events 24/7. That represents a lot of video and that represents a significant need for video infrastructure that can constantly be encoding that and distributing that to users all around the world so that might be like an example of a more traditional application and then you also see some kind of blockchain web3 centric apps experimenting with interesting models for Creator and musician monetization leveraging new web3 primitive so you see apps like the 402 which enable sort of token gated access to live concerts and live music experiences for followers of the artist or the musician or people who purchase a specific NFT or access to their loyalty program and that’s really exciting because that represents opportunities for blockchain and web3 to truly disrupt the traditional industry and create a better monetization experience for creators so we’re really excited about those types of kind of more experimental applications as well.
Mihai Grigore (4:51): Indeed it’s very encouraging to see demand growing and for us being able to quantify that in the number of minutes transcoded. Now moving forward to supply, we have the node operators that sustain the operations on the Livepeer Network and as we can see the fee revenue in Ethereum terms went up over the past quarter and also over the past year however the revenue in US Dollars went up over the past quarter but on a year over a year comparison it went down by 25% and similarly the revenue from staking rewards went down for node operators. So, the question is how is the Livepeer Network being able to sustain the operations over time given the current bear market and the down downturn in the price of Ethereum versus US dollar?
Doug Petkanics (6:07): Good question. So, a little bit of background the fees that apps pay into the Livepeer Network in order to use Livepeer infrastructure are paid in Ethereum, ETH. And then all the node operators all around the world that put their servers their GPUs on the Livepeer Network they get to basically market price what they want to charge for encoding video and they list that price in ETH right and so. As you mentioned in the lead up to the kind of the last question, I think Livepeer’s usage hit an all-time high and grew pretty significantly with over 40 million minutes of video streamed last quarter and so you would expect the fees paid in the network in ETH to go up and they did as you highlighted here. But you also mentioned the US dollar value of that ETH has decreased. That’s interesting it points both to sort of inefficiency in the market in that the nodes are willing to compete and provide the same services at lower cost. You’d expect that to happen over time as more and more operators come on the network maybe with better and more efficient hardware setups and they figure out how to lower their own costs so they can charge less. I think that’s the good side and you’d expect that to continue as there’s more and more scale and competition on the network. I think the little caveat I give you there might be a little kind of inefficiency in the market here as well where these node operators are not repricing the amount of ETH, they’re charging every single day or minute or hour as the price of ETH fluctuates in order to keep it at a constant dollar rate. If they were highly motivated and efficient, they could and they would but maybe the reason they don’t care so much is because they’re also earning kind of growing their ownership in the Livepeer Network through inflationary Livepeer token, so you earn the ETH that the users pay, and these node operators also earn the Livepeer token that the protocol mints out each round. And today that that Livepeer token represents kind of a much higher value and a much higher percentage of the node operators income than just the ETH that they’re competing to earn is and you know this actually also works in the network’s favor because it means you have to be less sensitive to the fluctuations and the price of ETH in the early days of the network before major scale and it can also subsidize their operations so they don’t have to charge as much for the encoding which makes the whole network cheaper for users which should attract more users and kind of kick off an exciting flywheel of lower cost more users more fees more competition lower prices and feedback around. So, I think those are some of the dynamics at play. Happy to dive into any of those areas if you’d like.
Stephanie Dunbar (9:00): Yeah just to touch on that a little bit more and talking specifically about the cost for end users. Something that we noticed this quarter, and this is also the first quarter that we were able to report on it with a bit more granular data, was that the cost for end users appears to have declined from 0.004 dollars per pixel front to 0.003. So, I have a couple thoughts in my head as to why that might be. I believe that it was over a period of three quarters that it went down but curious to hear from you what do you think is the catalyst for the decrease in price for end users for price per pixel.
Doug Petkanics (9:46): Cool yeah let me put that in a kind of a context that’s a little more digestible probably. So those prices are actually dollars per minute of video transcoded on the network and the .004 cents that’s equivalent to 24 cents per hour a video you send in. So, an app sends an hour of video it would be 24 cents and you mentioned it declined to three 0.003 that’s 18 cents an hour. So you see using the Livepeer Network current directly would currently cost about 18 cents per hour of live video that you’re sending in to be encoded again there’s some technical details that’s an HD video into three different renditions but what’s nice is you compare that to a service like Amazon Web Services which is $2.40 per hour or many other sort of traditional big company live streaming encoding services that are $3 to $6 per hour per live stream and you see how this current price of 18 cents per hour would be really competitive and cost disruptive it’s all over a 10x decrease relative to what you might pay for live transcoding elsewhere. And again, kind of as I addressed in the last question that’s not determined by Livepeer Studio, one of the biggest users on the network or any individual actor that’s just the market price that the node operators are charging, and I think your data is reflective of the aggregate kind of the average payments that they all received over the course of the quarter. And so, you actually see the benefits of open competitive marketplaces that take advantage of all this idle infrastructure that’s out there and how cost competitive it can be relative to the status quo which is really controlled by a couple big infrastructures: Amazon Web Services, Google Cloud and Microsoft Azure.
Mihai Grigore (11:44): Related to that this quarter we’ve seen the release of a video on-demand feature and being able to to play videos in a decentralized manner. I would like to tell us a little bit more on how we came up with the idea of creating this in the roadmap and actually having the release in Q4.
Doug Petkanics (12:11): Yeah video on-demand is a really exciting feature and it’s a big milestone for us and it’s one that’s highly needed. In the industry we call it video on-demand but you as a user of a social application can just think of it as an uploaded video that plays back like what you would watch on YouTube. The other kind of area video that Livepeer is always focused in is live streaming so that’s when you’re watching a live video being created in the moment versus something that already happened is recorded or created and uploaded and played back and as any user of any social application would notice if your app is successful you need both of these pieces. You need live streams, but you also need recordings of those live streams or maybe you just need uploadable video assets like what you see on TikTok and Snapchat and Instagram. And so Livepeer’s cost disruption and what it really enabled was very disruptive on the live streaming said, it was very hard to build a live video app at scale to operate your own infrastructure is very expensive and so Livepeer actually enabled startups and hackers and entrepreneurs to be able to do so affordably with live but until we could offer a video on demand product the offering wasn’t necessarily complete and the network wasn’t serving this whole other area of demand which is to be able to process uploaded videos as well, and so this is a really important feature that enables both of those pieces of the puzzle. It became more and more urgent as the as basically social apps begin to be built in web3 space and the blockchain connected matter on things like Lens protocol and Forecaster protocol and DESO protocol and Orbis in social that’s where every app needs both of these features and kind of Livepeer’s byproduct is the often the first thing that developers start building on.
Mihai Grigore (14:14): And speaking of social you also came up with a long thing NFTs could you give us some background here about what it is and how people can use it.
Doug Petkanics (14:28): Yeah NFTs as kind of an expression of content ownership and ownership of digital first canonical class that’s a really powerful and in the kind of the last wave of innovation around these we saw them move from kind of simple data-based constructs to image based. And you saw profile pics as a big use case, collectibles as a big use case. Often in kind of static images just like the way the first wave of the internet was all text based initially and sites like Wikipedia and bulletin boards and then the second wave came, and you had sites like Flickr which were image based and eventually you get to richer forms of media like YouTube and Twitch. Just like in this web3 space with NFTs as I mentioned you’re going to see more and more rich media but because we’re kind of firmly in that second image based wave a lot of the popular NFT platforms had limited video processing capabilities they were starting to experiment with video but they let you upload something like a 30 second clip or even a shorter a highlight of a you know a play and NBA Top Shots or something. This is cool this is a great first step but it limited the forms of creative expression that artists and creators could have so there was no way no easy way to mint NFTs from longer videos that were over 30 seconds or were over 100 megabytes so you couldn’t experiment with what’s an NFT as a music video look like? What does it mean to release a full short film or an episode or a full-length film as an NFT? As an ownable digital collectible that actually can play within popular NFT marketplaces there’s limitations on video processing infrastructure that these big players had and used and so Livepeer we created kind of a proof of concept and a capability the law and take NFT Mentor we dramatically expand the file size limit to 10 gigabytes which is hours and hours of HD video that can be uploaded processed played back instantly even stored on decentralized storage. I think there’s a lot of experimentation that apps can do with appealing to creators to create and distribute their content using this capability. Hopefully this sort of first long take NFT publisher demo that we put out there is a good showcase for them.
Stephanie Dunbar (17:00): I just want to say that having been reporting on Livepeer now for a year it’s been really amazing to see all the stuff that’s happened in a year. I think probably about a year from now there was a Livepeer transmitting network and the migration to Arbitrum and then after that it’s kind of been like an explosion of like new tools and NFT minters and increased file size limits and it seems it really feels as though the Livepeer core team and community are responding to the demand of the wider web3 social experience and just in general integrating even closer with the entire web3 community. I’d just love to hear from you, when was the catalyst that the shift occurred to grow and start to tackle more parts of the video stack and where do you see Livepeer, let’s say I mean in a year from now I can only imagine they developed the developments that will occur in that time?
Doug Petkanics (17:59): Yeah thanks for following all the progress over the last year and kind of recognizing all that’s been built. When you’re in it and you’re fighting these battles day to day it feels like you’re always just pushing things forward a little bit by a little bit. Like you said if you look back a year ago, Livepeer had a live streaming API and a live stream live transcoding infrastructure but was kind of barely plugged into a lot of the popular and growing apps that we’re experimenting with blockchain connected tech and now here we are kind of at the center of that movement as the video enabler and the video platform that all of these web3 social and creator apps are building on and that’s really exciting. I think the catalyst for this was twofold when his life here is six years old actually and we always had the mission to be the web3 video layer but you go back to when the network was launched almost five years ago and sort of the first developer go to market motions like hackathons and stuff we went to and four years ago no one was building kind of at the application layer no one was building social apps and creator apps in blockchain connected ways or the ones that that were were kind of very early experiments. Now you fast forward three or four years and you see the maturity of the space from a development stack perspective you see the maturity of the platforms that creator and social apps have to build upon as spurred on by scalable blockchains like Polygon and Arbitrum and Optimism and Solana you see really powerful tools for creator modernization like NFTs and all the Innovation that’s going around that space, and you see social frameworks like I mentioned Lens protocol and Forecaster and Orbis and DESO that make it easy for developers to get started with toolkits to build social apps and that’s so exciting to see it finally happening. Like we envisioned this future six years ago and now people are innovating at the application layer and that’s really where it’s aligned with our mission it’s aligned with the culture of everyone in the project and on the core team and so innovating and experimenting and meeting users’ needs in these categories has been really energizing and really rewarding for the team and looking forward like you mentioned where are we going to be a year from now where is I saw a question in the chat about the mission of Livepeer in 2023. I think it’s exciting to see all this momentum building in this space I still think we need to see the breakout killer social applications emerge that cross over to kind of mainstream awareness not because they’re a blockchain but because they enable better creator modernization, they better user ownership of the platforms that they create better curation of information for communities than what you’re getting from traditional social media today. I think this is the year that we’re going to we’re going to see that video is obviously a form of media that’s hugely important and impactful within these media centric applications and our mission is to make sure that these breakout apps the first examples of them are built on Livepeer and that should bring a lot of hopefully good to the good to the world and freeing us from some of the harms we’ve had from traditional social media and also good for the Livepeer network through its growing adoption.
Mihai Grigore (21:35): You have mentioned the killer apps and I remember reading your blog post from September 2022 on what’s going to be the next wave of decentralized social media killer apps. Indeed, we start seeing some early signs in terms of adoption and perhaps for you the question is do you have any update to your thesis. Did your understanding or your view of the world change in any way in the past three to four months?
Doug Petkanics (22:12): That’s an interesting question I hadn’t thought that my view of the world had changed and by the way my kind of thesis around here as I just articulated the building blocks are there the tools are there for experimentation around being able to take a stab at creating a breakout social application using web3 tools but we actually require product visionaries and hackers and entrepreneurs to take more of an innovative approach and not just try and recreate web2 social and recreate Twitter clones and Instagram clones and think that that is going to break out to the mainstream. I think it’s that you have to take advantage of what’s uniquely enabled by blockchain so that’s taking advantage of digital ownership and expression through NFTs permissionless crowdfunding through tokens and different distribution mechanisms taking advantage of user ownership and collective governance through down frameworks in order to create something new and new and innovative, so I think all of that definitely still applies and now is the now is the time to build. What’s happened in the last three months…I think the industry has faced some headwinds in terms of its perception in the mainstream due to the news focusing on frauds and financial blow-ups and failure of CeFi institutions to control their leverage and their risk management none of that stuff should affect builders in the web3 social space but I think if you put blockchain and crypto front and center of the top of your experience you’re going to face some skepticism or reticence and market acceptance. So what I would say is that that puts even more pressure on taking advantage of what’s uniquely enabled but making sure that you’re delivering it in a way that delivers real impact to the end users and they’re not there just for the speculation or the web3 and answer blockchain connect the nature of what you’re doing those are tools that you need to use in order to actually make real intact and create a great experience for users.
Stephanie Dunbar (24:22): I think something else that’s really cool about Livepeer that others or at least that I wasn’t aware of until this quarter is that the network can integrate even though the operator network is on Arbitrum the getting web3 enabled video or transported video from Livepeer can happen on any network so I think that’s really great because it’s a diagnostic to the chain and it can really scale to tons of different social apps and tons of different experiences that are available on different chains I saw that this quarter, Livepeer integrated with Aptos and I believe you can launch NFTs on Aptos that are powered by Livepeer. I’m curious to hear from you if there are more chains planned in the future and also what’s the engineering lift like for you know the life of your core team and community to enable that and maybe how many we’ll see moving forward.
Doug Petkanics (25:16): Yeah, I think yeah for anyone that’s been paying attention 2021/2022 it’s obvious that not all applications are going to be built on one single blockchain or one single ecosystem in fact there’s plenty of arguments for apps to have their own specific chains that all derive security in a roll-up or some fashion from a super secure L1. And so, it’s clear that social applications, creator applications, media applications are going to be built on all of these chains and Livepeer, the project should be an open video infrastructure that’s accessible to all of them. Excited to have an integration into Aptos. We’ve been running on Abitrum. We’ve, for a long time, have had an SDK for Polygon. I think Solana is a really exciting ecosystem we’ve tried to make headway with and support builders there as you find the right design partners. The recognition is that the what’s the blockchain use what’s the arbitrary in blockchain used for in the Livepeer Network. It’s used to incentivize and coordinate the supply side of the network, so the node operators use the Arbitrum blockchain to register on-chain coordinate and secure their work. There’s a staking protocol that routes work to the ones with enough security, it’s used to settle payments to node operators so those things can always exist on one blockchain Arbitrum. But to use the network specifically when it comes to video you’re actually not touching a blockchain at all you’re reading from one in your negotiating with these node operators and kind of an off-chain mechanism you’re sending video to them in an off-chain mechanism you’re sending payments to them in an off-chain mechanism that they can then use to cache on-chain later and so there’s a long technical roadmap that involves bridges for totally permissionless and native access to these chains. But what you often see right now is you see these other chains essentially using payment gateways like what’s offered at livepeer.studio just in order to fund sort of usage of the Livepeer Network and then they can use an SDK which is integrated with their chain to enable them to mint video NFTs that are minted in the protocols on their own chains. Access control is a really popular one so you control who watches a video based on what’s in their wallet are they holder of an NFT in a specific collection do they hold a token associated with the specific project do they hold a follower token for it on a social graph for a specific creator so if you’re building an app you can configure these permission controls based on on-chain logic in Aptos or on Arbitrum or on Polygon and certainly as you mentioned what’s the lift to add another integration it’s not too complex. I think it’s a couple week project from scoping to integrating to testing to having a lot of developer libraries or SDK out there.
Mihai Grigore (28:20): In terms of the experienced users we know that Livepeer is currently working on optimized features for viewing yeah short form content shorter videos in especially on capabilities that will enable seamless swiping can you tell us a little bit about what are some of the technical challenges in what does that require to make it happen in a decentralized manner.
Doug Petkanics (28:51): So there’s the Livepeer Network right which kind of specializes in transcoding and additional forms of compute today. And then there’s the Livepeer project which has the broader mission to be the open video infrastructure and that includes open-source software right and that software uses the network for transcoding but it provides additional capabilities around video that you need to use to build your application. We could go down a rabbit hole debating well does what are the incentives for people contributing to that software and does that drive additional usage to the Livepeer Network and I would argue it does but this this example you brought up short form video the types of videos that you see on TikTok where you have 10/15 second clips and the user wants to flip through them very quickly and they need to load seamlessly and they need to be ready for playback and smooth experience and whatnot. That’s very different than just saying buffering up and watching a two-hour movie on a video on demand or a GT platform. So this is where you get kind of into the open source software side so Livepeer is mobile SDKs and it’s player technology is open source and it needs to give developers the tools to be able to add these short form video experiences into their apps because users have come to expect them and because they create great engagement and this is why TikTok is the fastest growing app of all time and it’s why Instagram and Facebook have entered and YouTube have introduced their own version of shorts to essentially replicate that experience so web3 builders need that same capability the library open source software makes it easy for developers to get started with them and technically what’s that mean we’ll get too far into the details but it just means easy embeddable players that let you kind of cue up a playlist of what the next video, next video, the next video is it does all sorts of behind the scenes fetching and caching logic so that those videos the first few seconds of those videos are downloaded in advance and they’re queued up in memory and they’re ready to go so you don’t need to wait for buffering. The belief is as more users build on the Livepeer Network as you have more stakeholders in the network through Livepeer token they all have an incentive to make these tools better because it ultimately results in more demand and a more valuable Livepeer Network and so we’ll see them evolving over time. I think for builders that want to use and get started seeing how easy it is to add short form video to your application we just released a series of web3 video developer builder guides where people can go through a set of resources and quick start tutorials once focused on short form video that I just mentioned another is focused on the NFT publishing process so anything we can do to help walk builders through those those first steps is good to attract development.
Mihai Grigore (31:57): And how are these guides being received by the developer community so far?
Doug Petkanics (32:04): Positively they’re actually you can see them at builderguides.livepeer.org and they are a brand new project there’s going to be many of them there’s been two so far at least in the last two weeks the best kind of proxy for their reception has been they’ve been put out there and developers have the option to sign up for a workshop in which our team will walk through them and give you a quick start overview of how to use them and we’ve had nice 20 plus users sign up for the first couple of each of the first couple of workshops and show up in these are credible builders building real applications that are either launched or intend to launch soon and so again it’s exciting to see like the real substantive demand side interest and adoption of the utility of the network what it delivers as opposed to sort of just the supplied side token incentivized engagement like you see in the early days and in many types of projects.
Stephanie Dunbar (33:07): So just thinking about again the year that we’ve had at Livepeer or sorry we’ve had reporting on Livepeer so much has happened in these Builder guides I think are a really nice culmination of taking some of this open-source software and making it even easier for users to be able or for builders to be able to get involved and to build things. I’m super curious, lots of questions for you Doug. Are there any sneak peeks that you have any additional guides or things that builders can get excited about in Q1 of 2023?
Doug Petkanics (33:44): Yeah I think composability with a lot of the other elements of the kind of web3 builder stack is an exciting area and that includes kind of infrastructure composability with things like storage so an exciting kind of output of last quarter was that the Livepeer player can play back video that’s stored directly on IPFS or stored within Arweave through the bundler protocol and like given the content hash of an asset there the Livepeer player will automatically encode it into the different formats and Livepeer Studio we’ll cache it for quick delivery it actually makes video possible within These networks and that theme of like bringing video into the composable protocols that are being built in this this space is Livepeer’s opportunity it’s where Livepeer is really strong and I hope I hope we see that in the social protocols as well so making it easy even easier for builders building web through social apps to just assume that video will work with the Livepeer Network powering it under the under the hood so that that’ll be a big focus of the community it’s kind of building those right tools and integrations making composable all of these protocols in a way that a developer can solve a good experience and not have to interact with 12 different networks and learn 10 different SDKs and protocols in order to build the applications that they want.
Mihai Grigore (35:13): Indeed we’ve been recently analyzing storage protocols and we’ve seen also an uptick in in their usage in indeed given that Livepeer can now play videos stored in IPFS and Arweave will continue to drive more and more usage for the protocol in the in the coming quarters Doug any other thoughts that you want to leave the community with at the start of 2023.
Doug Petkanics (35:44): Just that Livepeer has been here for six years as I mentioned we’ve built through a number of tough crypto markets, and we’ve also built through a number of really exuberant crypto markets the projects always stayed very mission focused we’ve never strayed from being web3’s video infrastructure during all these years and we you know we’re not going anywhere and won’t stray far away either. We remain well capitalized. We’ll be building, growing the usage of the network, and making sure that as developers take advantage of everything and web3 we’re there to provide the video there.
Mihai Grigore (36:21): Alright I’d like to thank you both very, very much for this fantastic conversation and for our audience at Messari, we cover top protocols in in web3 on a quarterly basis to read our reports go to messari.io/research or on Twitter @Messaricrypto. Doug, what is the best way for the audience to read or listen or watch content about Livepeer.
Doug Petkanics (36:54): Yeah check us out at livepeer.org and you can follow us at Twitter on @Livepeer it’s where we’ll push out announcements about all the all the content for those who are interested in participating in the network, I definitely recommend joining our Discord it’s linked from livepeer.org. It’s a great community of node operators that operate the whole network and support one another and the project as a whole, so a great place to learn and be part of the mission.
Mihai Grigore (37:24): Excellent thank you so much! Bye everyone!
