Casey Winters advises MyArc founder Peter on marketplace sequencing: first deepen per-community network effects for retention, then layer in cross-creator discovery to boost take rates. He flags that matching and search are technically hard and need years of work, citing Whatnot, eBay, and Substack as models.
Peter, I saw you raise your hand there. You want to come on?
Thank you very much. And brilliant talk, Casey. Just for quick context, I'm Peter, one of the founders of MyArc, where we enable fitness creators to train their fans at scale. We've helped the supply side already make more revenue than they ever have in any of their social media, which is great. So lots of value between supply side and demand side, and also our supply side has acquired the demand side. very, very quickly as well. So kind of zero CAC, you know, for any consumer, which is great.
Yeah.
We're kind of operating in what I would say like a silo mode at the moment where right now, you know, when the fans come on, they, they don't see other creators. They just see that main creator they signed up via. But what we're noticing is that a lot of the consumers are starting to realize that there are other creators in our kind of MyARC universe and wanting to kind of go around and switch around. My question is kind of around, I feel like it might be too early to pull the trigger on having that type of like discovery or that kind of cross-networking and instead maybe focus on like the initial network effects that we can do on a per-community basis, if that makes sense.
Mm-hmm.
Because what we're seeing is that the social mechanics that we've kind of just released of community, et cetera, is getting, you know, retention and activation a lot higher. People are creating so much UGC that the creators in Reed put on their post and that's bringing in like organically more customers. So kind of diving a bit deeper into that, almost like a Strava-like interface within the experience and then going towards kind of marketplace. But yeah, that's kind of what we're thinking about right now. So currently operating in silo, we're seeing network effects within the kind of creator-to-fan kind of community. And, you know, wanting to know when to kind of, or maybe it's a bit too early to pull trigger on like that more traditional marketplace and discovery.
Yeah. Okay. So the question you have to work through to build the appropriate strategic sequence here is it sounds like you have traction in building out acquire creator, creator onboards community, community starts engaging with each other. That's a direct network effect business. And that's valuable. All direct network effects are valuable. And so the question is, do I lean into that? But there is some emergent behavior that maybe has high traction. So you could lean into that a lot more and that will create defensibility and faster growth, all things we love. But then there's also some emergent behavior that implies you could have cross-side network effects, which is cross-selling creators. And what's nice about that is maybe like the acquisition isn't as viral or anything as a direct network effect business, but the cross-sell will create higher take rates from suppliers because they feel like you're actually driving real demand to them versus just managing a community. So probably will make more money per creator. And you will probably see metrics over time that show the number of creators you get connected with, the higher the consumer side retention is. for, you know, the overall kind of product. So the LTV on the consumer side will go up. LTV on the consumer side is going to go up through the direct network effects too, but probably the cross-site network effects will be a little bit higher, assuming you're maintaining some balance of the community product. Probably makes sense to like achieve some milestone on the community side to feel like you're hitting initial returns and then sequence to better discovery unless you feel like the fact that you're not doing that means someone else is going to like that. That opportunity will go away, right? Like if you don't lean into it, it'll kind of evaporate for some reason. I don't think it would evaporate competitively. I don't think it's not like someone else is going to steal that opportunity. It's just more that people— it's generally been hard to balance historically, like community value prop and transactional value prop. Obviously Whatnot's doing a pretty good job at this, but historically it's been pretty hard, which is more on the community side. eBay is more on the transactional side. They were the first to have been able to blend the two quite well. So I think your prioritization probably sounds correct, but you're going to need to kind of reevaluate what are the gains from investing in building out the direct network effect? When does it feel appropriate to sequence to the cross-site network effect? So the other thing about the cross-site network effect I would keep in mind is getting good at matching is a pretty unique skill. And you talk to any marketplace founder and they'll just tell you how much they think their search functionality sucks and their discovery functionality sucks because there aren't good enough, there aren't enough good search engineers in the world. They're frankly all in China. So they're hard to get. And obviously machine learning and AI is propagating more and that stuff will become easier. But fundamentally just search is a hard problem. Matching is a hard problem. So you might want to get some work on that a little bit earlier, just knowing it's going to take a while to get good. Not distract from the core strategy, but if like, hey, if this is a call option, we're definitely calling in 3-year time horizon. It might take 3 years to get good at search and browse and connect. So now there are hacks to that, like take like the Substack model, right? Of like, well, if you follow one person, that person just recommends other people they think are good. That's usually like a good short-term hack, but not as much a game changer as you kind of owning the discovery model long-term.
