Building Respectful Online Communities

First of all, we're very clear that we're a community driven platform with community guidelines. So there is a set of expectations of how you show up. And one of the things actually through the course of this year we worked on was putting a neighbor pledge back up on the home screen. So the first time you come to nextdoor is either a new member or if you hadn't been back for a while, we now remind you with that neighbor pledge. A good neighbor, right? So, you know, be helpful, not hurtful, don't discriminate, no place for racism, et cetera. And by the way, we got there through a lot of work with Stanford Spark, so I don't know if you know Jennifer Everhart's work on bias and so on, but it's a great place where we often use the kind of psychology side of the house to come together with the products that we create. So that's number one, we are a guideline driven platform. So therefore we're clear on, on things that we will remove or even people we will remove if you're constantly breaking guidelines. Secondly, we are locally moderated. So this was a decision the founding team made long before I showed up. And I think it was a brilliant decision because it's based on the idea that really only with hyperlocal context can you know how a piece of content is actually showing up. There's no way that I think someone based on the west coast of the United States can have any sense for what my dad's neighborhood is female feeling about a post that say, might include religion, religious overtones. Is it tongue in cheek and jestful or is it actually hurtful? And so what should stay and what should go? And I think it's very hard to do that with just pure machines. And I'll come to that because we do also use ML. But the kind of the layer cake, bottom layer is local moderation. Now we've had a lot of work to do there. Even in 2020, we realized that a lot of our moderators were either overwhelmed or they didn't represent the diversity of their communities. So we actually rolled out a new program called the Community Reviewer. And I've been astounded. Like really fills me with hope for humanity. How many people want to moderate in their local community? We do use a layer of technology. There's kind of the easy stuff like profanity that machines can find, but over time, how do we train machines to know if a piece of content is going to get moderated? And then actually we inserted something called kindness reminder. So again, we're always trying to pull you back to, like, what is our purpose and our mission? Like, why would you join this community? Again, something Dr. Eberhardt worked with us on, which is this idea that if we know that what you're about to post is highly likely to be flagged as reported content, we tell you and we say, hey, remember, great communities are created with kindness. You might want to edit your response. And so instead of me writing back, tom, you moron, I might say something a little kind of like, tom, I really disagree with you and here's why. And again, that was premised on this idea that, first of all, people don't like to think that's in some future state. What they do is going to be reviewed and reviewed unfavorably. So we're playing on that human emotion, and just by popping up that interstitial, we slow you down. So we take you out of your kind of more biased dinosaur brain back up here, your frontal lobe, your thinking brain, where you're able to layer biases with education or shared experiences. So there's a lot of stuff that we do that is very different from other platforms. I think they're starting to follow somewhat. I've seen, for example, that pledge starting to get used in other places. And you know what? I'm fine with that. It's going to make us all a little bit nicer and kinder in an online setting. I think that's for the good of the world.