Nextdoor Holdings, Inc. (NXDR) Earnings Call Transcript & Summary

All right. So let me kick us off with our next fireside chat. It's my pleasure to have the team from Nextdoor here, Sarah Friar, CEO; Mike Doyle, CFO. Sarah, Mike, thanks so much for being part of the conference this year.

Thank you, Eric, for having us.

So I think, Sarah, maybe just take a step back. Obviously, you've been involved in the company for a number of years now and where the company is trying to go in terms of building a platform around local engagement and local community. Maybe talk about the journey you've been on in terms of the company and the platform and how it continues to evolve.

Sure. Thank you so much, Eric. So hopefully, everyone in the audience knows Nextdoor. We are the place to go to find out about everything that's going on around you locally. We often talk about it being the place where utility meets community. Today, we're over 80 million verified neighbors on the platform. We're a very engaged platform. So about 44 million WAU, weekly active users, that grew about 13% year-over-year in Q2. We're across 11 countries. Why do people come to Nextdoor? They come to give and get recommendations. So it's a great place to find something that you need locally. If you're a new mover, maybe you've just had a baby, whatever life event has happened, often Nextdoor is the place to turn to find that plumber, that hairdresser, that dog groomer, whatever it is that you need in that moment. Or maybe it's to actually give a recommendation back into the news feed to actually recognize a local business. Second big reason people come is our for sale and free platform, about $1 billion a month put on the platform. Right now, during kind of what's seen as more of a cost of living crisis, we've seen more and more neighbors coming to the platform to make some extra cash. It might be to find a side hustle, a job. It might be to sell something in for sale and free. About 20% of the items that we see up there are actually free, because people are actually trying to do something good for their community, or it might be that they're starting that business. And then the third reason people come is to give and get help. We were just talking about moments of crises, which are extreme moments of needing help, like if you're living in Florida and Hurricane Idalia is coming in at you, if you were in Maui, recently, like real tragedies, Nextdoor really rises in those moments. We see big engagement, a lot of new neighbors tend to join. But all the way down to kind of micro give and get help. It might be something as nice as maybe just checking in on an elderly neighbor, maybe helping you find the keys you lost on your run, but a lot of things that need local and presence. And then the final thing I'd say is we monetize through advertising. I'm sure we're going to come to this. But why would an advertiser come? Today, our revenue is broken up into about 1/3 small businesses, 1/3 mid-market customers and 1/3 larger enterprises. And why an advertiser comes to Nextdoor and not to another digital platform or social platform is, number one, we can prove it's real people at a real address, so we have location data, and we take the time to verify you. Second reason they come is very high intent audience, so people are getting stuff done, all the things I just talked about. So the ad -- a good ad will actually feel like a piece of content. I was demonstrating to an investor in one of our one-on-ones today and opened up my feed and my first ad was from One Medical, and it was an ad about getting your COVID and your flu shot. A neighbor could have posted that. And in this case, it was actually an advertiser, but it was excellent content. And the third reason people come is because we can do this hyperlocal but at a national or even a global scale sort of outreach. So if you are Home Depot, Walmart, maybe you've just refurbished a store, you don't want to blast that out nationally because it only matters to the folks who live within a radius of that store that makes sense. But we can allow you as that retailer to keep your national message, everyday low prices, but to then bring it down to a narrower context, like come see this at our refurbished store on 1800 Main Street. So I'll stop there, but that will give you a little bit of sense for who Nextdoor is and what we're trying to achieve.

Maybe I'll ask one more big picture question before we turn to sort of the elements of compounded growth inside the company. But in terms of how you think the vision for what the platform can turn into long term, what are you most excited about in terms of platform evolution and taking where you are today on either utility or use case and it continuing to evolve in the years ahead.

And so the great thing about Nextdoor is, first of all, everyone's a neighbor. So our TAM is global and should be absolutely massive. Second thing is we've proven, on the one hand, that local is hard, which might sound like a strange thing to say. Local is hard. It's hard to get density. It's hard to get it all activated. It's why we tend to be a bit of a Swiss Army knife, right? We've gotten asked in a lot of meetings today how are we different from an [ Insert ], Yelp, Angie's List, Thumbtack, Craigslist and so on? And the answer is, we're like them in many ways, but we're actually the two-sided network, right? We own the customer set. We have over 80 million consumers on the platform, and we're bringing them back weekly or daily even. So they know that we're the place to go locally. So we don't have to keep reacquiring them. It's actually why almost all of those names I mentioned are actually great advertisers on Nextdoor, too. So what we do is hard, but once we've built it, it has really high moat, and those usually make for really great business models over the long run. And then the final thing I'd say is I'm sure you've all been talking nonstop about AI all day. It feels like when I'm in a room with a group of investors, anyone in the tech industry right now, AI is everything. And what they had premised on is premised on data. When you're building LLMs, you want access to as much data as possible, particularly unique data sets. And I think Nextdoor is incredibly unique on this front. I think there's a ton of businesses and we can build on top of that data, but just don't forget the origin of our core competitive advantage right now, which is the local knowledge graph. It's pre-tagged data that can go into a model and really quickly train that model. We use it ourselves to create personalization in the feed. We're using it to do things like create an AI assistant for how you post. But I really think this is a unique asset that has yet to kind of show its full potential.

Interesting. Okay. You teased out a little bit of how the advertiser base has evolved in the last couple of years and a snapshot of what it looks like today. Talk a little bit about where you see the biggest opportunity sets as a company in terms of growing your advertiser base and how your go-to-market strategy around advertising, whether it's self-serve or agency relationships continues to sort of evolve.

Absolutely. So I gave you one parameter of our ad base, which is 1/3, 1/3, 1/3. 1/3 is SMB, it's truly local businesses. A third is mid-market, folks who probably have multiple locations, maybe they cross the state boundary and then a large national advertiser. If you look at how our overall go-to-market has evolved in the last several years, coming into 2022, right, we were on a high. Well, Q1 of '22, good quarter, Mike. We grew 50% year-over-year because we were just finding that we had these endemic verticals that were working, financial services, home services and real estate. Then we started to hit a little bit more. Inflation happened, and that really started to cut off a number of those verticals, financial services and to some degree, real estate. And then others started to really feel the pinch and they went to their variable spend first and pulled back. So what that did to us in having to kind of pivot, it caused us to get much broader. So today, we're a much more diverse advertiser base when it comes to verticals. So we have leaned into areas like health care, government. Actually, pets has been an interesting vertical because of the COVID boom. And dogs and cats, both a lot of them get lost, that's good for engagement, but they also need things like food and insurance and all of these crazy things that people now buy for their pets. So the verticals themselves have gotten much more diverse. So the good news, bad news is it's still a pretty volatile environment. However, the advertisers who are endemic to us have stuck around. We've given you that stat the last several quarters, if you kind of look back trailing 12 months, our top 50 advertisers, we've had a 90% retention rate. So they've stuck around. They just haven't spent as much because their own business is getting really squeezed. So I guess the optimist sitting on stage here has to say that, as you start to see maybe some relaxation in that, people begin to spend some more again, they're more confident in their business. We should actually see a resurgence of the core verticals, and we've added more diversity. The other thing we have done is added the diversity, though, on customer scale. And the big step there has been unlocking self-serve on our platform. So we have been spending a lot of time building out our ad tech stack. Think of it as a front end and a back end. On the front end, the key change is really enabling self-serve across a business of any size. So any net new advertiser to Nextdoor right now, whether you're a small mom-and-pop in the local community or whether you are a Publicis, you are an ad agency, you can build a self-serve ad on our platform. And again, that should open up a lot more demand for us. On the back end, we're starting to use our proprietary data. And so that's beginning to give us better targeting, better outcomes. Today, only SMBs are on our own back end. Mid-market and enterprise are still on Google Ad Manager, so we use someone else's stack. But we are now making inroads into that and are kind of moving at pace. So you kind of heard us on the Q2 call, we were pretty excited about the moves that we've made. So diversity in terms of verticals, diversity in terms of scale and diversity in terms of self-serve versus managed, which overall gives us a portfolio that feels much more resilient, less volatile today, but we do need the ad environment to also come back and spending to come back to really see that lift again in growth.

Sure. I would just add, and two other areas related to our acquisition of new advertisers, we have invested in our advertising agency relationships, which allows us to have a one-to-many voice. We have -- as we've diversified our verticals, we have really compelling case studies in many different industries. And that's an important tool to allow the advertising agencies to represent us to their client base and demonstrate really how to harness the power of the data on Nextdoor and create that hyperlocal targeting. So that's a way we've continued to expand our go-to-market strategy and provide diversification in the model as well. And then just one other area, too, on diversification is on campaign objectives. And so we've seen in a difficult environment where advertisers are more interested in true performance campaign spend where they can measure on the transaction basis or some actions to be taken, whether it's a click or a lead, we've been able to help them shift their spend and retain those advertisers and the revenue away from traditional brand awareness spend. We're still doing both. Today, we're probably 60% performance campaign spend and about 40% brand awareness, which is inverted from where we were probably 2 years ago, but it's great for us to demonstrate that range to advertisers, and we're happy to move with them.

Great. Okay. Sarah, on the last earnings call, the whole team talked a lot about the user funnel and how it continues to evolve as well as some optimism around the engagement trends you're starting to see. Talk a little bit about how you continue to think about user funnel evolving for the platform over time and how that's a backdrop for potential user growth as well as rising engagement and what that means for, not only user growth, but also the potential for ad impressions and supply on the platform.

Absolutely. So to kind of start with a visual, if I could paint it in front of you, think about places that Nextdoor can grow. We can grow the top of funnel which is brand-new verified neighbors coming in top of funnel. We use the term verified neighbor because we take that moment to ensure that it's a real name going alongside an address, so verified neighbor. Then we have to get you from VN, verified neighbor, down to become a weekly active user. From weekly active, we want you to then come more frequently, ideally to daily active, but we want you to go deeper in terms of depth. So that gets you from WAU to sessions, that's the number of times you're in the app. And then from sessions to impressions because that's how we make our money, ultimately, is we sell impressions. We might be selling a cost per click or cost per lead. But in the end, we have to serve a certain number of impressions to fulfill that outcome for an advertiser. So think about that as our funnel. If you go back to even 2019 going into the pandemic, we would have said we were not great at the engagement part of the funnel. We were pretty good at getting people in at the top. We had particularly leaned into paid marketing at the very top of funnel. We would send out invitations. I bet in this room for people who joined Nextdoor, you probably got a letter in the mail. It had a good CAC to LTV ratio, it's ROI positive, but still spends money at that top of funnel. But once we got you, you didn't necessarily stick around. So as we went through 2019 and 2020, we were really focused on that. And then frankly, with the pandemic, we've got so many people coming in at the top of funnel that we moved all of our focus to that midpoint of the funnel. So sitting here in 2022, '21, '22, we would have talked about kind of 5 things we were investing in. Number one was our notifications platform, which is really our code for ML. What you don't see is even though our headcount has stayed largely the same over the last couple of years as we've kind of really tried to be mindful of cost with a tougher top line, underneath that, we have really built out our ML team. It's about a 50-ish plus person team now. We've built the infrastructure for it. That was notifications to drive engagement. We added connections, which is a way to signal who do I know in this app, which also gives us more data. We built out our discover interface, which is the map-based interface that allows you to see what's happening around you. And then we built out just an easier simplified UI for the user, really oriented around getting you to take an action, ideally make a post. The fifth one, if you're counting, is our ad tech stack, which I'm going to take over here. So that was where our whole energy was focused outside of building ad tech. As we've gotten into '23, we said, okay, we're actually pretty good now in that engagement piece, right? Once we get you in the top of funnel, once you arrive at Nextdoor as a new neighbor, we're really good at getting you to stick around. So -- and we're very engaging. We're more engaging as a platform than Twitter, YouTube, Pinterest, like a lot of household brand names. And you stick around, 2 years in, our retention rates are kind of best-in-class. So 2023, we've gone back up to the top of the funnel and said, okay, we're going to put most of our energy there. And then we're also going to try to make sure that as you are coming through, you're going a little deeper in the feed. So that way, we can make sure that sessions keep outpacing WAU growth because that maintains our revenue growth. So if I go to top of funnel now, how are we getting new users? So it's really 3 things for nonpaid. And 90% of our new VNs are unpaid, they're organic. There's true word-of-mouth, and that's just a lot of we do our media, we show up in moments like we talked about disasters. So again, we can influence that, but it's a little harder to measure. The two big product areas that we are investing in, number one is invites and then the second is content sharing. In invites, we're taking all the learning we had in those paper invites we used to send into now the electric -- into bits and bytes and really thinking about how do we stimulate neighbors inviting neighbors. Some of it's easy. When we onboard you, we take you through an onboarding flow where we ask you to sync your context, and then we ask you to make an invite happen. We kind of treat it as a normal course of business. A lot of people do it. But one area that we're seeing really interesting green shoots in is getting companies or getting businesses to invite neighbors. And you might say, well, why would they do that or why didn't you do that before? Well, we looked at what we think is a very untapped opportunity. We have 4.2 million claimed business pages on Nextdoor. So we have 4.2 million people who could invite their contacts or their CRM system, the local plumber, the local hair salon, maybe up to even a cafe with multiple locations. Why would they do that? They do it because when they invite a customer and that customer lands in Nextdoor and saves them, which is they press a little heart, or recommends them, the likelihood that other neighbors will land on their page or react to their ad goes up significantly. The first goes up 5x, and on the ad, it's 6x. So it's not even 1%. It's literally 5 to 6x. So therefore, it's really in the interest of Matt's Plumber when he has a great interaction with Sarah to ask me, would I go to Nextdoor and recommend him and to actually be able to send me something that says, here's how you would do that. So that's a very interesting area of green shoots for us. On the content sharing side, there's 3 things we are doing. One is we share content out. Again, we changed the infrastructure of Nextdoor a few years ago where now we have this concept of any one content. But we're also investing in SEO and in partnerships. On the SEO side today, less than 5% of all of the new verified neighbors to Nextdoor come through SEO to give you kind of a comp, LinkedIn, TripAdvisor, that number is probably 70% to 75%. So we think this is a big area of untapped opportunity for us. It will take investment. It will not be a step change. It's going to be a gradual because you've got to slowly grab the Google algorithm so that you show up in search. But we think we have an IP, a domain that has high relevance and has high ownership of neighborhood, and we have a lot of dynamic information, which tends to be something the algorithm likes. So, so far, we're seeing interest in green shoots there. And then finally -- sorry, Eric, I took over your whole preso. We have been building APIs for partners so that we can both ingest partner content and also take Nextdoor content out onto partner sites. And that creates a flywheel of people who may not know Nextdoor get to see it. They see more of the content. They're asked to join. Bringing more partner content in creates more of the content flywheel, which really helps drive engagement. And partners are folks like The Weather Company we just talked about for moments like hurricanes and wildfires. GoFundMe, work really well in local communities. PawBoost is a site for lost pets. But what you start to see is we cannot build everything that is local. We may want to, but it's just not a same proposition. But we can partner into everything that happens around you locally, but make sure that we own the consumer. We own that neighbor that knows Nextdoor is always the place to come back to any time you want to do anything locally.

Got it. Okay.

Sorry. That was very long-winded.

No not at all. For the record, it's your presentation not my presentation. So I did want to stick with the theme and bring both Mike into this and you, Sarah. From a product side and then from a cost side, you talked earlier about AI and ML and in infusing multiple layers of the platform and what you're trying to build. Talk a little bit about what you're most excited about in terms of the product initiative side that you're building around AI and machine learning and how it might fund growth in the years ahead? And then to bring Mike into the conversation, how should we be thinking about some of the elements of cost that have to go behind those initiatives and how they could be dampening margins today but be accretive to margins as the growth comes through in forward years?

So I've already -- the good news is I've already talked about some of this. Our notifications platform that we've started really investing in, in 2022, that's AI in action. That's making sure that you get the right piece of content in front of the right neighbor at the right time so that they really engage. So you may get e-mails from Nextdoor, you might get an SMS, you might get a badge in the app. But it's our way of saying, hey, something just happened in your neighborhood something locally that we think would be of interest to you. And we've already seen that. Like the WAU growth that we've seen over the last couple of years, which has been in the 20%, 30%, a lot of that is driven by this investment that's going on. So it's not a new thing. It's been happening continuously. On the more kind of -- to get on the AI -- generative AI bandwagon, like excitement front, we have launched a couple of things that I think actually are really interesting and showing a lot of promise. So one is our assistant. When you start to type a post into Nextdoor, when you get above a certain character length, we'll suggest would you like the assistant to help rewrite this for you. So it's kind of like ChatGPT, but happening inside the Nextdoor app itself. It's trained on our data. It does use a ChatGPT API to do the rewrite, but it's trained in our data. And what we saw is we kind of just launched that and let every post kind of go through it if the user decided. What we've found is it works really well for recommendations, which is good because that's a chunk of what people do on Nextdoor. So if I'm either looking for something like a baker to make a cake for my daughter's birthday next week or I want to give a recommendation, it actually does perform better than the user writing it. And we know that because we obviously are looking at all the data after the post gets created to look at engagement, pre and post. Where it's still not quite as good is in the long tail. So where lots of iteration happening so that we think we can do it in more and more places. Because in the end, we want the neighbor to have a great experience. So you come to Nextdoor, I needed to get something done and I've got it done brilliantly because that will keep a user coming back forever, and they'll tell all their friends. But we also like that it drives more engagement on the platform because that drives more impressions, which drives more revenue. And then the one other area where we use it is around vitality. It's a little bit of a different use case. We have a prompt on Nextdoor that if you're writing something, we're already using ML to score what you're writing. And we know before you hit the go button that it's probably going to get flagged for moderation. So it's not a huge -- remember, it's a small -- overall less than 2% of all of our content gets flagged. But we know in advance. So why don't we try to stop you or change you? So today, we push up something called a kindness reminder, which says, hey, it looks like this might need to get rewritten, would you like to edit? Yes, it's a good idea. And we remind you kindness wins, we're like be kind. About 1/3 of people edit, which is good. But now we're not even making you edit, we're just offering you an alternate. And it comes back to philosophically how we think about Nextdoor that yes, we want to be part of all the really great things happening in neighborhoods, but we also understand that neighborhoods are gritty and often there's tough things where people disagree. But if we can help them disagree in a more constructive way, that just seems like that's ultimately a great thing, but it also helps tone on the platform. And when tone is better on the platform, engagement is better, right? It's an interesting thing because you'd graphed it. In the very short term, rougher content gets marginally more engagement, about 3% more. But over a 6-month period, gets about a 17% less engagement because people just don't like showing up and seeing hurtful content happening. So it's just an example of an investment. It doesn't really drive a ton more engagement. But I think over time, it just helps keeping to separate Nextdoor from maybe some of what we see going on in other parts of the Internet.

Got it.

Do you want to talk on investments?

Sure. If I can pick up there. I think the -- first I'll say I mean the good news is we have so many initiatives that we're excited about, whether Sarah explained today about our investment in the ad platform and now the opportunities to really leverage AI on the platform. Now the challenge is in the current environment, investors are obviously looking for improved profitability or a path to profitability for a business like ours in investment mode. So those are trade-offs -- tough trade-off decisions. And the reality is we are investing. We're investing in that platform. We're investing in AI. The good news is we've done that under the umbrella of our existing cost structure, have been very disciplined. And so as Sarah mentioned, the sort of rotation out or into AI, where we have quietly amassed an amazing team of almost 50 people on the AI team by shifting priorities. And while we've only had a very small growth in our headcount and low single-digit growth in operating expenses over the last many quarters. So there's not a big gulp moment of where we have just making a massive investment and waiting to see. We've made this part of our cost structure. And it's part of our current cost structure. And that's for us in a period where we're going to deliver EBITDA margin improvement. That's something we guided to for the full year. And so that gives us the envelope of what we have to invest, and we clearly have prioritized this as one of those areas.

Okay. Super interesting. Turning to the international business and maybe getting both your perspectives, how to think about the user opportunity internationally as well as monetization opportunity internationally, when you look at it from a penetration rate compared to the U.S.? It seems like there's a pretty long runway there to go after. So just understanding what the opportunity set is.

Sure. So just for the group here, I mean we're in 10 markets outside the U.S. The one where we're furthest advanced is in the U.K., where we're now in 1 in 4 households across the country, 1 in 3 in London and definitely at a scale that advertisers are interested and able to put to work campaigns and really get the results that they're after. So for us, we have a very similar growth model on how we're acquiring verified neighbors onto the platform. We use a combination of paid and organic marketing. We have a little bit heavier reliance or weight placed on paid marketing because we're just earlier in the development cycle. And so a little bit less brand awareness and a need to make sure that we're bringing new neighbors into the ecosystem and having a high rate of penetration that we know drives liquidity and engagement. So we're looking for ways to allocate investments into those markets and things like paid marketing so that we can continue to grow the impression base and grow monetization. So just to put some frame around it, about 20% of our verified neighbors are outside the U.S., but only 5% of our revenue is today. So there's a real opportunity as we get other markets to the scale that attracts advertiser interest to begin to monetize. We're doing so in a very small way in the Netherlands and in Canada. And as we migrate all of our advertisers over to our proprietary ad platform, then all of the markets outside the U.S. will have the same technology to be able to serve and measure ads and really drive the performance that we know advertisers need. We have a lot of latent interest in advertisers in those markets, particularly in Canada, where a lot of businesses in the U.S. have businesses in Canada and know how successful their spend has been here in this market. So we're excited to bring the international platform to their benefit in the next 12 months, really.

Great. Sarah, one of the themes that's been persistent in the last couple of years is a bit of the blurring of the lines between an ad dollar and a commerce dollar. And when, I think, people think of local, they think increasingly about commerce as a driver as much as advertising. How do you think about the commerce opportunity that sits in front of Nextdoor and whether you buy that, build that, partner on that? And you've taken some examples around driving commerce onto the platform. How should we think about that opportunity?

Yes. I mean I think Nextdoor when you built it, there's an incredible number of ways that you could then begin to monetize it. I do want to make sure we underscore for this audience that right now, we are focused on the ad tech stack, right? We are a small company, relatively speaking, and we have to get that transition to happen. That said, when it comes to social commerce or whatever term you want to use for it, what is it? Social commerce is the idea that I'm being influenced by someone that I trust. And there's no one more trusted in some ways than your neighbors, right? We do a lot of surveying of our platform for our bigger advertisers, and we know that something like 75% of all neighbors say, they've been influenced by another neighbor, 92% say they come to Nextdoor to look for recommendations. It's a little bit that kind of keeping up with the Jonases thing or some of this is just the information that you can't get anywhere else, right? I'm constantly in the market for an EV, but my main criteria is that I can get from here to Tahoe in it. And the car companies will tell me I can, but my neighbors will be like, not yet. Like you'll have to stop here to charge, right? And it's that sort of overlay that I think is really interesting, can we monetize that? On top of that, we have this whole marketplace called for sale and free. We see many ways over time to be able to broaden that, to put more neighbor made, neighbor might be barter in there. But there's so many other ways to kind of play that. Today, we only monetize it through an ad that we put into the feed. We actually took a step forward by putting Bing ads in there that are more contextually relevant, which kind of proves our point when we can use data to target, we could see much better outcome. So that's actually delightful to know that, that thesis does hold. But we see a lot of ways over time to build a much more vibrant commerce in the local community into that surface. But again, we'll start it. We want to get the ad platform right first, and then we'll move to the next thing and the next thing from there. But the good news is lots of way to monetize a base that already exists.

When you think about -- and maybe this is as much for Mike as you, Sarah. When you think about opportunities to prove out the operating leverage in the business over the longer term, is it really putting the ad manager and ad tech stack investments behind you and building scale on top of those investments that we should be watching for in the next couple of years that, that's really the unlock key to proving out scale and profitability leveraging?

Yes. I mean Mike should definitely dive in. But we've, I think, been quite clear that our path to being a profitable long-term sustainable, highly cash flow positive model is top line growth. As I said, you don't have to go back that far to a point in time when we were growing at 50% year-over-year. We're still not at some massive number where it's hard to get year-over-year growth. But with a combination of the base that we have today, a more open ad spending market and then the ability to utilize our own data, I think we're very upbeat just on that. And that's in some way, kind of what we've already built, can start to get you to look towards that $1 billion revenue number. On the ad piece specifically, think about it as today, right, we -- if an advertiser comes to us to get something done and we say it will take 100 impressions, we'll charge you $1 per impression. But we do that in a way that's very basic, right? We don't have amazing targeting under it because we're playing off with someone else's ad tech stack and we don't use our proprietary data. In a future state, we should be able to look at that and say, okay, we know exactly who to serve this ad to, to get to what you're looking for, advertisers, maybe it's a cost per lead. So we could serve you just 50 impressions and charge you $2 an impression. So same revenue, but we've only used half of our inventory. So it should almost feel like once we're on our own ad platform and able to leverage the data, it's almost like we get the step function in new inventory without having to do anything more than what we've already done. And then you layer on top of that the fact that we are still growing that I just talked about all the investments we're making to grow top-of-funnel, which will accrue to WAU, which will accrue to impressions and sessions. So it does feel like it's a forced multiplication, but it will take another year-ish to kind of finish out the build.

Understood.

Yes. And I'll just add, I mean it's -- as you rightly called out, I mean, it's really about top line growth to drive leverage. But we have a very attractive high-margin business, our gross margin is more than 80%. And our peers have EBITDA margins from 20% to 40%. And so it's really just important that we demonstrate our ability to get there. As Sarah mentioned, there's upside from doing more with the impressions that we already have through better targeting and driving yield up. We're making investments in -- have made investments in driving engagement, so that we have more impressions per WAU, so things like driving greater depth of session where we called out in our last call, we're up more than 24% year-to-date on impressions per session. It gets us more of the inventory the advertisers are after. And then the more recent focus on really driving top of the funnel growth is increasing the size of the unique audience. So what do advertisers want, they want to have a large, unique audience so they can do a lot of targeting and still end up with an attractively scaled audience at the end of that targeting. And they want to be able to put a lot of capital to work in a campaign to be able to justify the investment and optimizing on the platform, and that's really in the impression growth. And so that's really how we're going to end up driving that margin higher. I guess the last piece would be certainly the tight focus we've had on cost over the last 12 months and really being disciplined there, on keeping our OpEx growth very modest to deliver at least a couple of points of EBITDA margin improvement this year and being on a path to greater improvement in the years to come.

Sarah, just in the last minute or so, we have just your biggest priorities, whether it be investments in the platform or pieces of execution that you're the most focused on looking out over the next 12 months.

I think the biggest unlock for Nextdoor is that viral top-of-funnel growth, which is why we -- I talked about diversity of advertisers, how important that was to create resilience, but having diversity on growth. So it's not just one thing, but it's many things. When I joined, Nextdoor really grew because of word of mouth, which again, I said it's hard to measure and hard to count on, and paid marketing. We sent out invitation letters. Today, right, I articulated for you probably 10 different things. And each of those is getting better with every cycle. So that's -- my main focus is that because I fundamentally believe if you build it, you will absolutely be able to sell it. We already know that the ad platform is really high performing, and that's without any of the new ad tech that we're talking about. So if we can nail one thing, it's getting those new verified neighbors in at pace because once we get you, we keep you, we know how to keep making you more and more engaged, more people tends to lend to more engagement. And then we know how to build the ad tech stack around it to build a pretty big company.

All right. Well, please, everyone, join me in thanking Sarah and Mike for being part of the conference this year. Thank you.