How to Un-Poison a Well

Transcript

All right. Ah, and we're already on. All right. Hello everybody. So, my name is Maddie. I'm the head of founder experience for Polygon. And we're going to be talking about how to unpoison a well. And, uh, to start out, all of us here have some form of a community, might be developers, creators, contributors, users. And what I'd first like you to think about is when is the last time that your community made you proud? And hopefully there's many times. But then also when is the last time that your community made you nervous and which one of those outnumbers the other?

So hopefully uh so hopefully the first one is more than the other. But when it's the second one, this is about what you do about it. So we're going to be talking about what happens when the community turns against you. And so right now I'm at Polygon. Our team works with community builders, with uh developers, creators, and startup entrepreneurs. Anything anybody that's building cool stuff on Polygon, we help them do it better. And most recently, we're working on the Katana chain uh launching as well for the Katana DeFi chain. And before this, I was in gaming in the video game industry. And uh if you think crypto is toxic, oh boy, you have never uh released an update that

nerfed somebody's favorite character that they had spent several years of their life perfecting. So I have survived 4chan hate mobs. I've been put in charge of many communities that were left for dead and found out that there's a way to do that. There's a way to get them back on your side and working towards what you and the protocol are trying to do. And uh so what is a community? So what does this look like? And uh we've seen this before. You kind of you uh you know it when you see it, but specifically what you see happening is a bunch of nested feedback loops that get worse and worse and worse. They start ignoring what you say and then there's fewer people that are ignoring you which is surprisingly feels worse. And then

usually how people start is they say, "Ah, it's not that big a problem. It'll get better once we announce something new." or they start ignoring it. But then people start apologizing and they just apologize and you get in this uh apologetic loop and weirdly that doesn't work either. So then you start avoiding your community altogether and when you avoid your community somebody else is going to build it for you and that is much worse when those are the people that have turned toxic. And so the bad news is it's probably your fault. The other mistake that companies make is they say, "Ah, those people are assholes." Like, "It's all their fault that all of this went wrong.

If only we could just get them out of there, it would be better." And that's partly true, but the reason that the community as a whole uh went bad in the first place is probably because of decisions that you made months or even years ago. And these are decisions such as pulling out of the community and misaligned incentives that we'll get to in some of these upcoming slides. The good news is, and I believe this strongly, is that if you are the ones that screwed it up in the first place, that means you can get out of it. That means that uh you can follow a s that there's a series of plays that you can do in order to bring the system back to what you want it to be. And so, first we're going to we need to talk about the

building blocks of community. And uh ultimately, all community comes down to these three things. It's about the people that are there and who you're building with and who they attract to the community and how they decide to multiply themselves and it's about the community's job. Problems are good in this sense because the reason that people join communities is in order to solve a problem. It's in order to do a job and uh it's about the purpose that they lock in for about all right maybe they come to solve a problem but then how do they uh the crypto version of this has come for the tech and stay for the community and it turns out there's usually more than just the community from that. And if even one of these is off then all three will start going bad.

And then if all three have gone bad is where you start to get into that avoidant loop with the community mo. Here we go. So, the first one, people. And this is what most people think of of who is the I love how there's different emojis for every operating system. This uh and so what most people think of is the the people that are there. And the mechanics that happens is well, first that nobody wants to hang out with So the one psychological mechanism that is important to remember here is you might have a hundred people in the community and you might have five people that are uh that are the naysayers that are the loudmouths. If you don't have a good rule system in

place that can identify and show them the door then that's all that the new people see because they're the loudest. Most people in a community are lurkers. I think it's 90% will never post anything and that's okay. But if the uh 10% are all saying everything is terrible even though you've been making it better. If they are only focusing on the negative, then the people that come in say, "Oh, this place is for people that are upset about stuff. I'm not upset about stuff, so I'm going to go hang out somewhere else." And the answer for this is to shrink. So, make the rules clear. We'll have some uh tips for that in a moment. But to make the rules really clear and shrink down to a very small community, 150 people is your max,

but you probably want like five five or 10 when you're doing this reset. And uh you want problems. What? So something that drove me a bit baddy. So uh I was at a metaverse company before hopping full-on into crypto. And uh this was specifically a Roblox competitor for uh game developers. And uh so I talked to probably hundreds of game developers and asked them what are your wants, hopes, needs. And most of them said, "I want a community. Can you help me build a community?" And then I said, "Great.

What would you like your community to do?" And they said, "Hang out. I would like my community to hang out together." And uh that's a terrible thing to build a community around because people don't hang out together. They hang out while doing other things. You go for a walk with a friend, you have a couple beers, maybe you play video games together, but uh without having that reason for being for gathering folks together and the problem that they're trying to solve, it just becomes idle chitchat that then people start uh start showing themselves the door. And one of my favorite uh community principles was from one of the creators of Battlelet. So, this was an early online gaming service and he said, "If your wiki is not growing, it's dying." And part of this is because you

need new content that's coming out, but also it's because real people in real life die. So, there's no such thing as perfect retention in a community. Uh they might have kids, they might leave out of it. There's going to be some reason that the community decays. And the problem is actually the mechanism by which you get new people to be joining the community. One of the best examples for this I've seen is the curly hair subreddit where this was just a random subreddit that uh like other hair communities was mostly posting photos of your hair which I'm here for it. But then somebody on the community discovered really like the optimal chemistry for curly hair and they rebuilt the entire community specifically around that chemistry and

product recommendations and it became one of the largest hair care communities on Reddit and has a whole industry that came from it all because they were solving the problem of frizzy hair. And uh the purpose is why they stay and it's the point in the distance that you point to. And the crypto version of this that's really important to watch out for is building around price. This is the uh the crypto version of hanging out where great we all want to root for a price that means it's doing well. But even if you're really really good at it, most of price is not in your control because it's it's a random walk. Things could go bad in the economy. And then with people

rooting for it, they are happy when the number is up. They're angry when the number is down because there are some people that are angry when it's down. Nobody wants to hang out with and so people leave. And then you get to people are angry even when the number is up just because they remember times in the past when it wasn't. So uh the fix for that is you have to have some kind of distant point that people are working towards that is not out of their control. A quick quick story about that is uh so prior to crypto while I was in the gaming industry I ran an esports studio for a while and this was based on uh early community building we did for a Nintendo game called Super Smash

Brothers Melee many years ago and we built a competitive community where people were trying to beat one another for money and we found out that if you get people in a room to be trying to beat each other it can turn hostile like they start trying to withhold their strategies and try to prevent other people from winning. But we made it about being the best and finding the best. So that was then about going to tournaments. It was about practicing. It was about finding who is the best in the region and funding them to go to nationals. So we were able to have that same competitive spirit but built around something that was much much much more sustainable.

So these next four slides are ultimately what you can do about it. We're taking those previous three things, the building blocks, and each of these can uh help you reset and help you start building on top of it. And we're going to get a little bit more practical here with some programs that you can steal. And uh ultimately it usually so it ultimately starts out with shrinking it down and starting small. Doesn't matter if you have 10,000 uh community members if most of them have left. Signups is less important than active users. And a way to do that is pick less than 150, but again probably start with your five to 10 best users. And the five to 10 users that you most want to clone over

and over and over. You get those five or 10 people, you get them into a small private space together. And if you think about hosting a dinner party, when people are first arriving to the dinner party, you hang out in the kitchen because it's this small space that you have something to do, solve the problem of uh you haven't gotten a drink yet, and then people hang out there, start building up culture, and then it's this nice space when more people start arriving. So, you need usually this looks like create some kind of dedicated channel. This also works because if it's a private channel only, they have access to these five to 10 users feel special.

And get rid of the chaff. Get rid of the gosh. If you have a Discord channel, well, first I have questions for why you're still using Discord, but uh if you have a Discord server with channels that haven't had a post for uh six plus months, archive it. Nobody's going to miss it. It is worse to see all of those empty rooms and roles that people aren't paying attention to. and uh you get it down to that 5 to 10 and it's with that group that you drop the drop the new rules. This is when you start showing them what the purpose is and it starts very mechanically with actual rules. You need to say this is okay and this is not okay. Usually there will be a couple of obvious things that you see like oh we

shouldn't have people tearing each other down for having a bad day in this other spot. You do need to give them something positive to do. Otherwise, it just shows that you're banning all of the fun stuff, but ideally you're saying uh so like Flickr way back in the web one days is the ones that first did this where they said here's five things you should do, five things that you shouldn't. Now go take photos and share them with one another. So you get the rules, you post it, you put it on the put it on the forum, on the Discord, and as soon as those five to 10 people start following it, you reward the hell out of them. You put them on pedestals, you engage with them, uh you promote them on Twitter, on any channels that you have. And a flip side of this is uh you have to be really

careful who you pay attention to because other people are watching that as a signal. So it is very tempting for the loud mouths, the problem children, to spend all of your time trying to fix them. And that's problematic for two reasons. One is there's not enough time. But then uh everybody else is looking at that and says, "Oh, if I just act up, I'll get attention from the rock stars, from the people in the community, from uh the ones that are building the protocol." So they start following the example of the ones that are getting the attention. Flip that on its head, post the rules, support the people who are uh actually driving the culture. And something else will start to happen when you do this, which is people that were

borderline or even pretty healthy members will also start to leave. And this is a little bit different from the ones that you're showing the door where if you did a good job of resetting the rules, it is necessarily different than the rules that were there before. This means that you will have some long-standing users that say, "This is not for me anymore. I'm going to go do something else." and uh they might post a thread about it. They might say, "I'm leaving. Goodbye." Uh the thing to do is hear them, talk to them, understand it, but let them leave. Don't try to stop them. It's because you have now like if you are changing it and saying that this is what we're now about. Some people aren't going to want to do that. So,

it's a sign that you are sending a strong signal. In a community, we call this the uh don't forget your toothbrush phenomenon. Just say don't forget your toothbrush. We're here. uh we're here if you want to come back and then the good ones come back. They see that it's still there in like three six months and you start to see them popping up again. A uh a version of this that we did for Katana was so Katana is the chain that is launching today. I think mainet is going live in a couple of hours. Thank you.

And uh uh we wanted to do the same thing for DeFi there where the community was first about early access, about sharing secrets, about access to this new group of information, but then gathering everybody around make DeFi fun again that yeah, you can search for yield, yeah, you can try to exploit the system and all these ways, but find ways that it's fun and then share that with other people because that's ultimately how you grow it. And ah the problems my my favorite uh my other favorite story about problems is from a guy named Rich Millington who's a super OG community builder writes a great newsletter if a great newsletter

if you're looking for one and he told a story about a time he took over a teachers community and this teachers community had tens of thousands of members in it and completely dead. It was the number one community by far, but no activity whatsoever. And so then he did a crazy thing, which is he started talking to them. He had one-on-one conversations with all of the power users, past and present, with the people who had just joined and really just listening for, hey, what made you join? What are you working on in your life now? What are the biggest problems that you're facing? What are your goals, hopes, dreams, and fears? And the number one answer that he that all of them came back to was time. There's not enough

time. And for teachers, there's even less time because they're grading papers. They're dealing with parents. They're building out the lesson plan. And uh they're just always trying to find more ways to make their time work. So he rebuilt the whole community around that. This was a this is an old school vulletin forum, which is near and dear to my heart. But he got rid of the channels that nobody was talking in. build the whole thing around your lessons plans that same save save time, your tricks that let you get all this done faster and activity started shooting through the roof. Suddenly all the mechanisms were working again that the existing users were posting content that the uh that the old users were coming back and they were starting to see new faces that stuck around once

they found it. There is actually one other thing there which is get a win. So like get like build programs around uh the problems that you surface but uh with that group of five to 10 people with that group of 150 people tie back the successes to what they said. And so if they at some of times this is even just by DMs, but if somebody uh gives you a piece of feedback that then you incorporate into the product, send them a note saying you uh incorporated it into the product, but then tell your social team to then thank that person if they're down for it or thank the community members because this goes back into progressing that feedback loop.

And once it's working, do more of it. And uh uh this is another one that sounds obvious but was like revvely for me when I realized it is a good community program tends to have some of these same building blocks but then you can actually repeat those building blocks across lots of different programs. Find a new five to 10 people find a new problem to solve and find a new output. The other secret to this is then you also need to talk to your protocol to your company about the problems they are trying to solve. These should in the end be the same problems.

they should be aligned towards solving the same thing. Otherwise, another community might be a better fit. But this is how you align community programs to the business goals of the company itself. And uh an example of this is from notion. I found this on a notion thread like two years ago. I can't find it again. If anybody knows of where this is written up, please send it to me. But notion said, "Here's a list of all of our community programs." And then listed like 10 of them. They had a community program for making templates. They had a community program for colleges, university students, for uh spreading with inside a business and giving people resources for that, for making other kinds of content. And then after each of these, they listed the incentives and

bonuses for each one. And they were just all the same. They were access to the team, early looks at roadmap, and special perks, and maybe a paid version of the product. So you can actually take those and then just uh what they did was they said we want more templates and people on university campuses and word of mouth and companies and plugged these same systems into each of those and then we're able to run all of those because they had so much overlap. Oh man. And we got there in 20 minutes. So thank you y'all.

And I'm told we do have time for questions and I would love to answer any. Hi Masha from Uniswap Labs. I'm curious like as a community like builder and steward, what are signs when you have more push back internally? Um what are ways to justify or advocate for your work? Um, and maybe like if there are some signs um that you should be aware of where maybe the community isn't being valued as much. Yeah. And what to do.

Another very classic community question. And you see this in a number of ways for how to present to executives, how to get budget, how to get resources. That usually comes down to the alignment question. And so some uh some community builders will sometimes build up a community that is not plugged into an output that the company cares about. And if it is something the company cares about, then it should be more of a straight shot to communicate in the language that they're trying to speak. Um so something that didn't make it into this deck, but that helps me for communicating this is to really emphasize to the people that I'm talking to that what community is really good at is engagement. If you're trying to get

community to do acquisition or retention, it's actually quite bad at that. Like uh so product is going to be 10 times more effective at getting people to stick around and marketing is going to be 10 times more effective at getting new people into the funnel and the way but the reason that it sometimes looks like these can affect that is because they are amplifying the inputs into it. So, uh, what we just did with Katana was we knew we wanted to drive word of mouth and Nick on our team started hosting a, uh, Twitter space dedicated to Kaido where he was just trying to give something useful to these teams and to these creators and help them grow their personal brand. And, uh,

once we started seeing that that was working, then it was talking to the marketing team and product team for what their goals were. And it was also that word of mouth and that outcome. So now we have this pool of really amazing creators that are feeding into some of those same awareness metrics that the marketing team is trying to handle. Um metrics can help. So just uh anytime you can abstract away from uh using phrases like a lot and uh metaphorical language for and I've worked with community builders that will like literally say they have torches of torches and pitchforks and everybody hates us which is only rarely true but uh if you can say 50 people in the last two months

have raised this issue and uh 95% of them turned out even if that's not perfectly aligned It helps add some specificity there that can help build out the problem. And does that answer your question? Hey. Um, so as you were talking, I heard you say like, you know, you should be using Discord, right? Um, uh, my question really is like where should community be or what's the role of social media versus Telegram groups versus whatever else is out there?

I like that. And I have tried so many times to kill Discord and it's not completely gone, but it's mostly that these days it's not really great at anything, but there's some stuff that it's better than most other things at. Um, I find that Discord is best for an enduser community that uh you have some uh that you don't have those really focused ways that you want to focus them yet. where basically like I think of discord as you need to have something so that they don't go somewhere else that somebody like can turn towards another protocol or the like there uh it goes back to talking to the users so uh actually another katana example here we were working on the discord community as

we started moving up to launch and every builder that we talked to said uh oh yeah we don't use discord anymore we use telegram and then we resisted it tried to make it work in Discord But we kept on getting that feedback. So we said, "All right, screw it." Put everybody into a Telegram group. And it actually worked. Like it worked way better because we focus it down and didn't try to do everything with it. Uh but then we found out that the uh the rituals we wanted to build like having spaces and live audiences and um giving alpha to the community did have hooks in Telegram that they could use. Uh but yeah, it goes back to where is your community most likely to go in search of solving

that problem and you should probably own it. This is another spot that reasonable people disagree with me, but uh the I'm a strong believer that the company or protocol should ultimately have the keys for the community that they're in the community space that they're investing in. um to the extent that they can like with decentralized things like Farcast are totally fine if they don't have that but it should be a place where you can have rules enforce rules and show that it's a clean well-lighted space. Anyone else?

Thank you. Um, I was wondering if you could expand a little bit on the example with the curly hair subreddit. Uh, I find that you may have some more insights there. I I saw you cheering for the curly hair example. I was so happy. Yeah, I love the whole thing. Yeah, there's some So, if you d if you dive into it, there's some really practical things that they did that immediately make it made it clear that it's what the community was about. And uh once they realized that there is a whole lot of chemistry to this uh they started creating a list of all of the ingredients to avoid and then categorized I think four different hair types and they started building tools around it. started out with a list on

the wiki and this is like uh folks gathering together to bring all of this information that started showing up on search engines. But then the top members of the community started building tools on top of that. This is where it starts to look more like Devril where you're supporting people that support the rest of the community. And so, oh, one of them built a, and these days you can do it with chat GPT, but they built a drop a list of ingredients in the text box. It'll tell you if it's curly hair, uh, curly curly girl hair friendly uh, friendly or not. And then they also did this in a way that engaged the community where they would have monthly recommendation threads. So, what did you find? I think they call them grails. And

so a ritual they build around is finding that grail product because everybody's hair is a little bit different. And so they would have the monthly thread that then people started posting their grails into and then other people would start talking about that. And then it went through all the other life cycles of a community where uh it started getting successful. So then blogs are making money off of it and then they're you end up having a fork in the road where where the community goes is different than what got them there. But there are ways to navigate that also. So, by like having that uh like that clear end point that you're shooting for, and I'm definitely curious if you saw any others in there, like any others that pop up.

All right, anyone else? All right, friends. Thank you so much. Y'all should stay for Bruno who is going on next to talking about similar community building principles especially for how to get developers excited building some stuff that everybody is happy together with. [Applause]