Onchain Messaging: A Fireside Chat with XMTP CEO Shane Mac

Shane, so great to have you here today. Thanks so much for being with us. I think the founders in our Founder Fellowship have a lot to learn from you in terms of your experience, um, getting adoption for XMTP as a messaging protocol, um, the decisions that you've made in terms of what to include in the protocol, as well as the general importance of Web3 messaging, um, and maybe even some implications for how they think about boosting engagement and retention for their own applications that they're working on. But before we jump into questions, I wanted to give a brief introduction for Shane. Um, so Shane is the CEO and co-founder of XMTP, which is building an open protocol and network for secure Web3 messaging. Prior to XMTP, Shane was building in the Web2 space. He was the founder of Assist, which built the first messaging platform for Facebook Messenger, Apple Business Chat, Google Business Messaging, and SMS. For those of you guys who remember all of those different messaging platforms and how businesses were leveraging them, that company was acquired in 2019. And before that, Shane co-founded and bootstrapped a company called Squared Away, which provides remote EAs, um, power powered by military spouses. Um, probably there's a few folks in the room who've either worked with Squared Away or have heard about it. Um, and that business recently passed $25 million in revenue led by his co-founder. And Shane is based in Nashville. So Shane, thanks so much for being here today.

It's, uh, really great to be here. Thanks for having me.

So first of all, I think it would be great to kick off with just what is crypto-native communication? What is Web3 messaging? Can you just sort of define that for us?

Yeah, for sure. I, I, I look at it as there's a new identity, which is a wallet address, which enables actually the ability to own infinite number of identities. And in that world, how can you make sure that every identity can reach each other? And so for, for the way that I, I kind of looked at it and the way I thought about should we start XNTP, in the past, messaging and communication usually follows a new identity mechanism happening. And so whether it was a website leading to emails, whether it was phone numbers leading to SMS, the question for me was really, is this going to be a new identity? And then how is that going to play out? And, you know, I think it's still playing out, but it's, it's clear that the ability to reach a wallet address that you otherwise could not reach through any other mechanism needed a new way to do it. And it's kind of how the company started. Robert Leshner, the founder of Compound Labs, he was talking to us about 3 years ago and he is like, I have $11 billion in a smart contract and I can see every single person's address, but I can't reach 95% of them. And that was really where it all started.

I think that's a great segue to doing the demo and like showing people what the experience looks like and, and putting it in a more visual way. I think that's the best way to get people to—

Does everyone here have a Coinbase Wallet?

I think most people do because we asked them to download Coinbase Wallet yesterday. So amazing. Folks have it on their phones.

Why don't we just do this? Why don't I just open Coinbase Wallet? Why don't I go in my inbox and why don't I tell everyone to scan this, which will drop you right into a message with me. So if you scan this QR code, it should take you right into the inbox and it should be right into the inbox to message me. If anyone needs it again, I'll put it back up.

Shane, do folks need to first scan the QR code to activate?

It should activate it actually while it's doing it. So it, it should drop, drop you right in there and it'll activate it if you don't have messaging turned on, but it'll also go direct to it if you do. Hey Nick. Hi Lee. Whoa, I'm getting all kinds of messages. So what's cool here is If I go out of Coinbase Wallet and I go into actually another application called Converse, I actually use Converse as my main messaging app, but I also use Coinbase Wallet. And what's cool is the messages are fully portable. So I think what's completely new here is you own your conversations like you own your assets. The application and your private key is actually the thing that decrypts and takes your communication with you. So you're not locked into any ecosystem. So if I go back into Coinbase Wallet, I can see, I'll see all of your messages coming in here. I can do Face ID, I can open up and I, I can see all of the, all of your messages. So everyone who just started messaging me, if you've never messaged me before, it starts out as a message request. So it's kind of cool. I can accept or decline it and say, say Lee here, I can go in and see conversation with her. I got a message right here. I like your emoji. Cool. The unique thing that Coinbase did is if you hit the plus button, you can actually do payments. So I don't know who this person is, but maybe it's your lucky day and I'll send you $50. and now someone in this group just got $50 from me and the message receipt actually goes into the message feed. And so that's what we call standard content types and the ability for content types to be able to be interoperable where now you can have secure messaging globally, but also because of crypto's killer use case is the ability to do USDC and payments globally. And now you can actually do that altogether in one. And the coolest thing is that the message is verified by the private key. So the person you're talking to and the message itself, the actual message object is provable that it's from the person's ID or identity they're using and the person who owns what they say they own. And so one of the biggest problems in all of Web3 is The scams happen on Discord and Twitter DMs and all these other platforms. 'Cause you can't prove that the person that you're saying, hey, this is my address, just trust that it's me. I get anxiety every time I send money. But what's really cool is that you have the ability to do payments now and messaging together in one seamless experience. The no more sending 8 cents and 10 cents as like a little verification of a wire. The message actually is the verification because it's talking from the address that you're actually the same as you're sending the money to. And what's cool is then if you go into Converse, all of those messages show up. And so I actually have all my conversations, you know, hey there, what's up? And so I can use any application I want. And the magic part of this is yesterday a company named Duffle pitched me. And what's cool is they don't have to, they don't have to rebuild a social network to start. And so for developers, there's no more cold start. So they built this whole crazy experience and all this stuff and these new features and all these new ways to show content types, et cetera. But I opened the app and my entire history was there. And so just thinking about no more cold starts as a problem, I think is a fascinating kind of look into where this all goes in the future.

I think that's a, a, a really, um, interesting and like provocative point, which is that as a developer, like integrating with XMTP basically gets you past the cold start problem if you're building any kind of, um, social use case or, or any kind of application that could benefit from messaging. Um, I think it, it would be great to hear. From your perspective what you think some of— to hear deeper thinking on like what are the benefits for developers to XMTP when you go into these conversations? What are you pitching? What is really the value add that XMTP brings to an application developer? And how are you able to get those first few integrations in place?

I think there's two things to think about here. One is why were so many messaging platforms started that didn't work in this space? And that question to me is, I think people were looking at it as, can you message this new wallet address identity and use that as a way to send messages and then building it in a silo? And they're thinking, oh, I'll use this public identity and then I'll use that as the, as the thing, and then anyone who uses my app can actually message each other using that identity. And that played out over and over again. And, you know, it's not like what we're doing is actually this crazy novel new thing. I think it's a unique approach to what makes it different. And the problem wasn't, can we use a wallet address identity to send message? The problem is, is there a ubiquitous network where everyone adopts so that you can make sure that all the messages get received? And that latter problem is a harder problem, but I think it's the one we're solving. And so to solve that, you need to figure out where are the shelling points that have the most reach with the largest inboxes that can also get all of the identities onboarded. So therefore you're solving the cold start problem so that any developer who plugs in can now reach everybody. And I think for us, that was going for wallets. And focusing on wallets are the center of the universe. How do we incentivize, work with, get them to be investors and be on the cap table, be a token holder in the future? How do we make sure that the entire world of wallets has a reason to work with us as well as has a reason to work with other wallets? And I think thinking about why should Coinbase do this themselves? I mean, that was where we started 2 years ago, and obviously that's the first place everyone starts. And I said, Would MetaMask ever build on top of Coinbase messaging? Probably not. But now you have this public identifier and I'm going to be like, hey, I sent you a message, Lee, just FYI, it's to your wallet address. And every single app has a different inbox. It really just doesn't work. And it kind of through throughout all of the times that this has been tried, that's been a thing. So I think it was very intentional why we didn't build an application and we just focused on the network and protocol layer to try to highlight why The problem wasn't, can you use this address as an identity to deliver a message? The problem is, can this address be reached in a ubiquitous way across the entire ecosystem? And then you have to work on, and I think what we're still working on is aligning the incentives and the consent protocol so that all those applications get value in the ecosystem itself. But really, I think focusing on the core wallets as the shelling point and then aligning thousands of developers to that vision so that if we do have a ubiquitous network and there's a way that every address can reach each other, that every developer gets a ton of value. But if we end up in 25 different silos, no one has any value there. So that's the, that's the, you know, the game we're playing. And I think it's a long road, but it's going in the right direction.

It's definitely a really fascinating shift from the Web2 messaging world where all of a user's messages were extremely siloed to that particular platform that offered messaging, and there was no kind of open identity system or no concept of messaging one canonical profile. It was all coupled with a particular product or a platform, and, and obviously you're chasing a very different vision now that is newly possible. Um, I think another interesting topic that we have been discussing as a cohort is the question of, um, getting adoption for a protocol and building the first product on top of it as a protocol developer. So a few weeks ago we hosted Hayden Adams from Uniswap, um, and talked about product versus protocol and basically how Uniswap was also the developer of the first product on top of the Uniswap protocol and how that was an important strategy for them to get liquidity into the protocol itself. I would love to hear your, your thinking on this topic. And did you guys ever consider building a messaging application yourself to seed the network and to get user adoption for the network itself?

The organization you have to build, and I think with startups, it's all about the sequencing of what you do and the constraints by which you operate. And getting that wrong is how you die. And so, well, if We had Coinbase Wallet but didn't have an app. What would happen? Like, that's what everyone's asking for. If we did have our own app but didn't have Coinbase Wallet, no one seems to care. Yet we're spending tons of resources and now staffing up a go-to-market team to think about consumers using our product and that whole thing versus our customers, developers versus our customers, users. And that was really the deciding factor for me. And we had this debate. Who is our customer? We're only going to pick one. If it's both, we're not, it's just not an answer. And at the end of the day, I think our customer was developers. And from where I come from, my whole reason for doing this company was I spent 15 years building on top of Web2 platforms, getting screwed over by API changes and then building an app to beat me with an app. And I was like, my whole reason for being here is actually to create a better experience and a better incentive for developers to be able to build with the protocol that all works together. And the incentives actually move the value and money to developers, not to us. And like, I still believe that. So we killed it, and I think it was the greatest decision we did.

Love that. Continuing to pull on that thread of developers being your key customer, could you share a little bit about the developers who are building on XMTP today? What are the types of applications they're building? And what are they seeing so far in terms of how users are using messaging?

We actually see it as just two pillars. So we have inboxes and we have platforms. So inboxes are a place for daily use, the common denominator touchpoint with XMTB. This would be Coinbase Wallet, this would be Converse, this would be all the Lens DMs, this would be, other, other wallets is really kind of the focus. And then platforms are Web3-centric dApps that would need re-engagement or a place to communicate to wallet addresses that could otherwise not reach. Think Snapshot, think Paragraph just launched last week, think OpenSea, Lens, et cetera. And I think there's two kind of pillars of use cases inside of those, B2C and peer-to-peer. On the B2C side, I actually feel like that's the real problem today. That can drive a lot of value for both sides as long as it's an opt-in and a consent-based relationship. And so Paragraph launched last week. Now you can do all of your newsletters. It can be delivered via wallet. You can get that in Coinbase Wallet or any other application on XMTP, and then the notification actually comes there and that you consented with your wallet on the Paragraph site. So seeing creators and fan, like in, and writers and artists starting to actually do one-to-many kind of announcements, but also they turn into one-to-one relationships, I think is the killer use case. There's a company called Holder.xyz, there's a company called Absolute Labs, and the Converse team just built broadcast.converse.xyz. So if you go there, it allows any creator in the entire space that's ever launched an NFT or minted anything to an audience creator can actually reach that audience now in a more targeted and a direct way to, to have a conversations and build relationships. I, I think a lot about Snapshot use case. I think governance and doing that in a really meaningful way where the end-to-end loop, you can send a notification about governance, you can actually governance, you can actually vote inside of the wallet and then never have to go back to the site and completing that loop is super powerful. I think everyone will have a wallet newsletter or like an interoperable newsletter in the future. So a subscribe button on every single site. We're thinking a lot of like, imagine the ERC kind of update to the NFT, 7, I think it's 720. Imagine there's a sign or a checkbox or another signature that is the opt-in to communication. So imagine that on the, on the heels of every mint, you are actually giving consent or not to letting people message you and having that relationship and with that creator. And so those are all like, I think really cool use cases on the peer-to-peer side. I really believe what Coinbase kind of cracked, which I showed, and the ability to do secure communication with global payments for free is a holy grail use case that is just much bigger than people realize. Because when I see WhatsApp try to do payments for a decade, but the payments rails are so complex to be able to pay people across borders, and then I can't pay someone on Venmo in Canada or even Mexico in the US, let alone, you know, anywhere in Europe. Crypto's killer use case is actually global payments. And so to see L2s and Coinbase support free gasless payments mixed with secure communication that's verified. You know, this is like Western Union, Venmo, WhatsApp territory, and those are just huge businesses and opportunities. So I think those are cool. I would say the, the discovery stuff is pretty amazing too. The Lens, Farcaster, using any identity or social graph to be able to help, you know, and discover who to message. That's probably something I could have showed at the beginning. Converse, if you go into Converse, the app, they actually do an incredible job at doing discovery from on-chain data, PoApps, social graphs, et cetera. And I think building and discovering your graph and your people based on the communities you're already a part of or the identities you have is, is a huge use case.

Yeah, that's very cool. I think we're like shifting from this world in which our social connections had been originated from offline sources into online sources. And I think the next evolution of that is that a lot of our social connections originate from on-chain sources. And So being able to discover people based on shared on-chain activity or, um, PoApps that you have or things that you've minted is definitely a very cool use case for XMTP. Um, we're gonna transition to audience questions soon, but before we do that, I think it would be great to get your advice for a lot of the really early stage builders in here on general go-to-market strategies. And insights, especially, you know, given that a lot of the founders here are just launching their products or about to launch and hoping to gain traction. It would be great to hear your advice on just general go-to-market in Web3.

It's an interesting question. I feel like my, all of my advice would be the opposite of all of Web3's advice. 'Cause Web3's advice is like, make sure you have the token, the token's a growth model and all these things. Like when I came into the space, you know, I'm from messaging and building secure messaging for almost 15 years. My co-founder Matt, way more crypto. So he's very much, last company got acquired by Kraken, started building on Ethereum and right when it launched, like huge in the space. And I think we kind of defaulted away from that of what I've learned in Web2, and I think we're learning here is One thing I think about from growth is truly knowing who your customer is and not building for both. So when I heard, but yeah, we need an app, but we should also do the protocol and we should also do this and think about the Chrome extension. And it was really like hard for me to understand, like, is the end user, are we building a team for go-to-market is actually about monthly active users and it's all about the application metrics and getting people to open the app daily or is the customer developers friction to adopt SDKs? What is the failover rate? Do they have interaction success with who they're trying to reach, etc. So I think picking that was very important, which goes into my like main philosophy of think smaller to build bigger. I really think that this space specifically needs really, really great products for a lot of really big problems and take like just everything happening around like smart contract wallets or account abstraction and the, just the onboarding SDKs. Like there's a lot of companies in that space, but that to me feels like an obvious, just like amazing problem to be solved and focusing that specific on really this thing solves a developer's problem like completely. It's 10x better than having to deal with what wallets are on your phone, the WalletConnect connection, I can now support email and SMS. Consumers don't have to deal with it. And like that specific of a problem is a huge idea. And so I always just try to keep reducing the surface area. So then my go-to-market is so targeted. I think go-to-market follows actually getting that right. And a lot of this space feels like they forgot the product principles and then they tried to go to market with like philosophy around token. And I don't know if that's right or wrong, you know, and we chose to go the other way. But I feel like if you do have product market fit and you do have adoption, and adoption comes from really specific products, having good constraints, being able to make sure that every single person really loves it, then you get to launch a token. For me, I'm so excited about that because we can have all of our like ducks in a row, not launch the token to then find If I can find the product market fit, and maybe that's completely counter to this entire space, but I think I'm really excited about that approach.

Yeah, I think it's the approach that we advocate for as well, finding product market fit first and building that community.

And this is where I go to market's tricky because go to market, like, get users, get growth, whatever. For me, go to market is always the opposite. It's, you end up lying to yourself as a founder by thinking the next feature is gonna be the one that grows versus editing, reducing, or removing the features you have that aren't growing and really continuing to edit or change or throw out. And sometimes look at experience, like throwing out our entire app level and redoing the go-to-market to a whole different org. We started over. And the fear in this space I felt a lot at the protocol level was, but we can never change it. And I would be like, God, I'd be like to my co-founder, be like, hey, we're a year in. If the answer is start over, but we have a year of experience and we can go 6 months and do it again, I'm fine. But I think there's this fear I've felt in this space around like, oh my gosh, we can't, it's a smart contract, we can't unwind this. A blockchain lasts forever, like all this shit. And I'm like, just shut it down and start over. But that experience I had from the past where when you try to morph things or you try to like add to it, or I'd always ask founders when I was an investor, my, my trick question would be like, what are you going to build next? And the founders that actually told me a bunch of features, I knew they didn't have it figured out. But the ones who were like actually were trying to actually get this one thing over here to work, it's kind of close. No one's really— we're going to edit it down, we're going to remove a bunch of shit. We might delete this app, we might try it again, we might do a different idea. That was actually a signal that they had the loss of, the loss aversion strength, which is the hardest thing. It's twice as hard to remove something versus add it. And that's the thing to actually like lean into because that fear and that feeling is the thing that really will help you get to a simpler thing. And that I've always, the people are so scared of loss aversion of, we, we probably spent, you know, Lee's an investor, so I probably shouldn't say this. We probably spent over a million dollars building these apps, probably more than than that. I mean, we built up a 6-person app team. I had to let 2 people go. We got rid of our head of engineering 'cause he was app-centric. Like, it's been not an easy shift, but I think it was really the right shift and the editing and the constraints. And I say as a leader, reducing your surface area actually leads to go-to-market because if not, you're trying to go to market with a bunch of things. And then the hardest thing is that if you don't have a thing, you can lie to yourself that 6 things are kind of working and you really get fucked up when you like have 6 things that kind of work. Because everyone's like, oh, well, this, they all, they all kind of work. And that's where you end up in like spirals that can go downhill.

That resonates a lot. Simplicity, paring things down, finding what, what actually works rather than overcomplicating it. Um, let's switch over to audience questions. So the chat has a lot of questions in it.

Sorry, I feel like I just got preachy there. I didn't mean to.

No, I, I think a lot of founders resonated with what you just said.

That's the hardest thing. And so I empathize with all of you as well, and I struggle with it every day too.

Yeah. So I see a couple of questions in here about the business model for XMTP and how you plan to monetize. So let's tackle that one first.

Totally. I think we're getting, we're making a ton of progress here and I'm super excited about it. It also goes into the other question I saw, consent and spam. Coming from the messaging world and, you know, from the last, 15 years of my career, there's no bigger problem that I'm excited to focus on. And I think what's interesting is when you have a default public identity, you need a different mechanism for which you do consent. So email and SMS were supposed to be default private. All of our data got resold in the world. So ICANN spam walls got created to like try to protect us, but really it's just black boxes trying to like protect senders. Social actually invented a public identity, but supposed to be private messaging. So, you know, Facebook owns WhatsApp and Messenger and Instagram DMs and Twitter DMs, and you kind of have a social graph so you can do a lot of things. In this world, ENS names, Farcaster names, wallet addresses, it's default public. So I think you actually need a default opt-in network. So the biggest thing we're excited about is a proof of consent model. Which is how the world works today. If I'm a brand, I want to put a subscribe button on my website, an opt-in button. I want to put an opt-in button underneath the ads on Instagram. So if you're running ads on Instagram right now, you can do a send to WhatsApp and you can do a WhatsApp Messenger button. We would want one to say opt-in, and now you can get messages via XMTP on any client you use. And I think the opt-in being opt-in by default. So people control their inboxes and you, this kind of would change everything where you can't just message anybody. The only reason everyone here could message me is because you scan my QR code. So the QR code is actually an opt-in mechanism. I think you can do a lot of things around links. So users could actually manage tons of links. Oh yeah, make one for anyone with this NFT can message me. So now if they tried to message you, that actually would be a consent mechanism. And the second piece of this is putting that all in a smart contract publicly. So every XNTP identity lives on a chain, and that's public. The consent mechanism of can Lee's wallet message Shane's wallet is a private attestation on that identity, whether or not that identity is yes or no. And I think that to me is like a fundamental good for the world because that consent today lives in the sending applications of the SMS tools, the SMS marketing tools, support tools, spam engines, and all the government runs around trying to like make sure everyone got those numbers the same. And if you want to sign up on a website for a new SMS tool, they're like, hey, it's going to take a week to turn it on because we need to go look at the language you use. And are these your real phone numbers? Do you have consent to do this? But no one really knows. So I think that's the most valuable thing we can do. And then the business model is very simple. It's brands stake money to be able to reach the people that have opted in. It's the exact model of WhatsApp. And if you have a million people for nike.eth, Nike is happy to stake $50,000 in the protocol to be able to reach their most valuable relationships. And I think the innovation there is actually if you do staking, at scale, you can actually pay the ecosystem and fund it on yield. And so we're thinking a lot about how to actually do that, where then brands, if they ever wanted to leave, one, they can take their network with them 'cause it's a wallet addresses, but two is they can actually just pull their stake and they didn't have to lose any money versus today they pour all their money into SMS and email marketing and it's just gone. So we're thinking a lot about it there, but the money will be made in one-to-many sends. And I think the innovation is actually Uh, the building a, building a protocol of consent.

Yeah, that's really interesting. I don't wanna take us down a rabbit hole, but I imagine that it will be difficult to ascertain who's a brand versus who is a consumer versus a prosumer versus like an influencer who's like borderline user, borderline brand. Like I think all of those things probably need to get teased out with this particular business model.

I don't know, cuz I've been in it so long that I've realized you can't figure that out. When Messenger tried to launch this, they actually tried to add tags for like, oh, this is a marketing message, this is a support message, this is a brand, this is a person. And all it did is actually everyone just figured out how to hack the system to build to be something else. They're like, no, no, I'm just a person sending a bunch of messages or whatever. They're like, this isn't marketing, this is news. And they had all these different variations. I think if you have an opt-in network by default where even lead to message chain can't happen unless I set my inbox so that anyone with a Lynx handle can default opt-in, right? Can automatically, can automatically message me. Or if you scan my QR code or if I send you a share link or if my website says, You know, I have a button and actually you can click that. But if you actually don't differentiate between all the different types, you're just basically saying everyone has to get an opt-in, even one-to-one. Could it slow the growth of the network? Maybe. But unless everyone actually wants the control and they can create easy, someone could say, let everyone message me, right? And that would just actually turn it where any message that tried to message them, would turn green as a, as a consent because they, they set that, but it's a user-controlled setting.

That makes sense. Okay, so we're running up on time. This has been super informative and interesting. Shane, thank you so much for coming in to chat with the fellows. Um, definitely gives us a lot of food for thought. I think you also got a lot of new users for, um, the messaging experience inside of Coinbase Wallet and, and all of the other places where XMTP is supported. So let's keep messaging there. And I guess, yeah, if you have other questions now, you can reach Shane directly inside of Coinbase Wallet or any Lens application or anywhere where XMTP is integrated. So I'll volunteer Shane to continue answering Q&A in that messaging experience. So thank you, Shane, so much for being here.