 Oh, all right. It's kind of nice hearing my voice on a nice, loud sound system. Well, all right. So welcome to the growth crash course. I will be teaching you how to do growth, whether you're an early stage startup or you are an early stage startup, early or later stage startup. Moving on. So I'll go really quickly over where I worked. So Corembase, Angelus, Strava, Haven, Airbnb. Haven was a tiny seven-person startup. And I think when you think about growth, it's really nice to know exactly how large the companies you are working with. So what worked at Haven, when it was a seven-person startup, is very different from what I was working at Airbnb. Working at Coinbase, where I'm working in Web 3 in crypto, is very different from when I worked at Strava. So the nice thing about growth is that once you start learning about growth, you can take those learnings and move it somewhere else and generalize it and kind of adapt it to fit the correct niche. So I'm a product manager. So obviously, I had some goals with this talk. And the goal is basically this. One, I want to make sure that everybody is a pretty solid foundation of how to do growth across multiple channels. And the second one was that I wanted to make sure that I provided concrete examples and insights every single step of the way. I think this is super important because when you talk to people about growth, it's really easy to go very, very high level. And most people aren't able to provide very, very deep insights onto what's worked for them and what doesn't work for them. So hopefully, you get that. So the very first thing that I'm going to tackle, this is really boring, what is growth? So all companies grow, but startups are defined by growth. And I like to use growth with a capital G. And I define it by the process in which you are very intentional about growing. This is the things that you do in order to get more people onto your platform to get them to re-engage, to get them to perform very, very high intent things. So the very first thing I like to actually, so this is the agenda. I'm going to talk about funnel optimization. Then I'm going to go to acquisition, virality, and then activation and engagement. And I like to start with funnel optimization because this is a very fundamental growth concept. And I think everybody here has heard of it. Everybody's almost done it. But it's very applicable to all areas of your product, not just growth. And it's a really nice way in order to kind of dip your feet into growth if you've never done it before. And I like to break it down to three steps. First one, break down your product into discrete steps. This is the user journey. This is how people are going through things. The second one is discovering the inefficiencies. And the third one is to fix them. Step one and step three are super, super easy. It's not difficult in order to break it down to steps. And then it's very easy to fix them outside the engineering problems. But the hard part actually is figuring out where the inefficiencies are. So in order to do that, my favorite place to start working on growth, if you've never done it before, is sign up and onboarding. Because sign up and onboarding is already a natural funnel. There are steps that your user has to go through in order to onboard. And there's a very, very, very clear goal that they need to go through. So on top of that, every user that goes through onboarding is high intent. If they're on onboarding, they want to get to your product. And you want them to get to the end, which means that any change you make is going to have a very, very, very high impact. And the funny thing is is that for as key as this is, almost every company I've ever worked with, sign up and onboarding is very under-optimized. And it kind of makes sense because when you're a smaller company, or sorry, if you're an earlier-stage startup, you're trying to build things as quickly as possible. So you're going to build out your onboarding funnel, and then you're going to move on to building out the actual product. And when you're a larger company, the onboarding funnel is probably one of the oldest things in your code base, and it's kind of a pain in the butt in order to actually fix. So because of that, it's usually really complex in order to fix, but it has really, really high return. So with that, I'm going to walk you through how we did it at AngelList. So these slides are really small, but step one, if you remember, is to break everything down into discrete steps. So we break this down into pre-onboarding, onboarding, and post-onboarding. So for pre-onboarding, you hit the home page. Fine, you go, hit, sign up, and then you enter the onboarding process. Then this is a seven-step process in order for you to actually finish. So what do we do next? Try to figure out what the inefficiencies actually are, in which case, the very first thing we do is you look at amplitude. So we have gone ahead and built out the instrumentation in order to see what steps users are going through and where the actual drop-offs are. So from this step, we know that our onboarding funnel had a 30% conversion rate, and it was a pretty big drop in between some steps. Then after that, I'm going to steal Holly Lou's saying where data kind of lights up the room, but it doesn't tell you where to move it. So even if you know where things are bad, you don't exactly know what you need to fix, in which case, then you start to rely on different things. So we used full story in order to look at user sessions. You can see how users are clicking around, what they're doing, and then on top of that, we also did some UXR interviews in order to figure out in order to get actual feedback on how customers were using the product. And what we found out was that at the very top of the page, you can see we have this very, very, very vague bar that just kind of fills up. And because it just fills up, you don't know what steps you're going through, you don't know how much time you need, you don't know how close you are to finishing it. Users are very confused about the title. They're confused about a lot of things just from looking at all the feedback that we had. So this is what we ended up doing. So what we did was we went ahead and made it a little bit more clear about what steps you need to go through. We went ahead and made the steps a little bit more concrete, and we added some nice little messages to say, save and continue, because that means that they are not losing where they are. But the most important thing that we did, and it's not on the slides, I apologize, is that we reduced the number of fields that you actually need to go and input. So I think one of the things that we found out was that we asked for your location four times. So that's like, where do you work? Where do you live? San Francisco, San Francisco, San Francisco. So what we did is we just consolidated it into one single field. And that way, the less field the user has to fill out, the less chance that you introduce friction, and they don't make it through. The last thing that we did was actually very interesting in that you can go ahead. So you need to upload your resume at one point. We took the resume, we moved it to the very end of the funnel, because when you need to upload a resume, that's a very, very, very high amount of friction. You have to go find it on your desktop and upload it. And by moving it to the very end, the interesting thing is that because you go through all these steps, answer all of these questions, is that you're so invested in finishing the onboarding process that you're going to go ahead and upload the resume anyways. Whereas if you had the resume up first, and you're like, ah, this is too much work. I'm not going to do it. So by doing all of that, we increase conversion rate from 30% to 60%. And the more important thing is, is not only did more people convert, they actually started doing more job searches. And this is super important, because not only are users finishing the funnel, they're performing something else that is very, very intentful. All right, now we're going to move on to acquisition, in which case, I will be talking about SEO and paid marketing. Everybody talked a little bit about conversion and optimization. So we're going to move on a little bit from there. So SEO is really nice, because it's free user acquisition. Whereas with paid marketing and performance marketing, you have to go and pay for it, right? And in times of where it's a bear market or you're hitting a recession, SEO gets a little bit more attention, just because it's free. You don't need to pay for it. And for SEO to work as an acquisition strategy, it's a function of both quantity and quality. You want to have a lot of pages, and the pages need to be very, very high quality. By high quality, there's an SEO saying, content is king. The more content you have on the page, the more successful your pages are going to be in the Google search results. And I usually break this down into two different steps, either programmatic or contents. We're going to go with programmatic first. So programmatic basically means let's go and generate as many decent pages as possible. They don't be great pages, but they don't be great pages. They could be kind of OK pages. But you're really good for quantity over quality, because what you're trying to do is you're trying to generate as many pages as possible to bring as many keywords as possible. And this is a very scalable strategy. And it minimizes the amount of risk that you have to invest in SEO. But it's not an option for every single company. So you have to think about whether or not this is going to work for you. So to look at Airbnb, Airbnb has a lot of city pages, right? Places in San Francisco, places to stay in Helsinki, places to stay for every single city in the world. Zillow, lots of house pages, Instacart, a lot of product pages, Zapier, a lot of integration pages. You can see how this would work out. And as an example, I worked with this company called Spot Angels. And what they do is they provide parking information for specific cities. And what we end up doing is we created these pages so they can make for places to park near the Statue of Liberty, places to park near Golden Gate Park. And by creating all these pages, they went in three months from 60 visitors a month to 100,000. And the only thing that's limiting them is the amount of information that they have in order to create more pages. But why don't we take a look at somebody that does not a great job at it? So if you look at Ramp, they have these pages called Travel and Expense Management in whatever industry, right? They have a bunch of these. This is the first page. T and E, which means nothing to me, for logistics companies. And then the second one is for prop tech companies. Look at these pages. Nothing is really different about them except maybe they use a different image, and the header is a little bit different. And what ends up happening is Google has no idea what these pages are. So if you Google management software for real estate companies, it shows management solution for e-commerce. Google has no idea what these pages are. These are terrible SEO pages. And their programmatic strategy isn't particularly working. But let's move on to content. So content is fun because you've got to be very, very targeted on the keywords that you want to rank for. So if you are a skin care company, you can go and I want to write a page for skin care routine, in which case you can go attack for very specific keywords. But it's really high risk because it doesn't always work. And you're putting a lot of effort into creating very, very high quality content. So if you do want to pursue content, this is the general rule of thumb. It's that you probably want to start with blog posts and landing pages. I usually try to start off with trying to rank for keywords. I get about 250 searches a month just because the competition is going to be lower. And usually lower volume keywords are going to have a much higher user intent. So they will convert more. And then once you get that going, then you can target bigger keywords after 1,000 plus, 2,000 plus, et cetera. But for all of SEO results, you should see them in about one to three months. I think a lot of SEO experts will say, oh, it'll take like six months to a year. They're lying to you or they're not that great at SEO. Moving on, paid marketing. Paid marketing is fun because it's very, very easy to spin up, which also means it's super easy to lose money. So if you're going to do this, you need to be very careful and very thoughtful about your strategy. And this is also used to the first acquisition channel for a lot of different startups. So I would say that if you want to do paid marketing, you should do this when you make money or you are charging. Because paid marketing is entirely about how efficiently you are able to spend. Your customer acquisition cost cannot outweigh the LTV. If it costs more money to acquire a customer than the customer's worth, then you're just going to be bleeding cash. So you need to know these two numbers in order for you to actually do this efficiently. So for small startups that want to try doing this, my biggest piece of advice is try out different channels because you don't know what works depending on where you are. So Google is always a really, really good place to start. It's always fantastic. This is how they make all their money. Facebook has been great but recently not so good. Twitter's never been good, especially with the recent news. Maybe it will be even worse. So I would maybe not do Twitter. And I find influencer marketing to be pretty interesting now. But really, what you're trying to do when you're doing performance marketing, you need to know exactly where your customers are and go and try and buy ads there. So if you are a, let's say, more developer API-focused company, buying ads on Stack Overflow is probably going to be much more efficient than anything you do on Google, Facebook, Twitter, or influencer marketing. And then for the larger companies out there, this is a study that we did at Airbnb. So this is a quote from our head of performance marketing. If you guys don't know what branded searches are, it's searches that have your brand name in it, right? So for Airbnb, Airbnb, San Francisco. Airbnb, Helsinki. People are typing in your branding keyword in order to navigate around your website. And what Airbnb ended up finding out was that it's not super incremental. It's about 15% incremental that you're getting. But it was ridiculously efficient, which means that it was really cheap in order to buy the keywords and then get users onto your website and actually bring in money. So there's always this fight on whether or not branded search conflicts with SEO. And what we found in Airbnb was that it was really fine in order to have both of them coexisting. And my favorite thing to talk about is virality just because it's pretty deep. So virality can go to referrals, word of mouth, social sharing, things like that. Mostly be focusing on referrals for this talk. And it's fun because it's one of the most nuanced channels that you can actually go into because you can grow your user base super, super quickly and they will advocate for you, which is really, really cool. But it needs to be very thoughtfully created because a lot of bad things could happen because it kind of relates back to the customer acquisition costs that I was talking about earlier, where if it costs you more in order to get a referral than your users are generating, then you'll just start blading money. Yeah, I actually just talked about this. I think the only thing I wanna say here is probably the last bullet point, copying other successful companies may not be the best idea because referrals is really weird in that your company's positioning, what your customers are like is gonna have a really, really big difference on what an effective referral strategy is going to be. So I'm gonna talk about Robinhood because Robinhood is probably the best at virality period that I've worked with and I really like their referral program. So the first step for Robinhood, or I guess for anything, is that when you have a referral program, you need to figure out how a user's gonna get there. If a user doesn't know where to go, they're not gonna be able to actually refer anybody. So if you look in the upper right, it says Rewards, that's how you get to their referral program. But the interesting thing about this is that I think like three or four years ago, with Uber, Lyft, and Airbnb to substitute for their referral program, we didn't use Rewards, it just said free $15 and users would click on it in order to discover the actual referral. So anyways, this is one thing that you can optimize is how do users go ahead and actually find out how to get to your referral program. The next step once you get there, this is fun, is this page says, invite friends, pick your free piece of stock. And when you think about referrals, there's three different ways to look at it. It's either selfish, altruistic, or quid pro pro, which is you scratch my back, whatever. But the way that Robin does this, it's very, very selfish because what it says is that you get the free piece of stock, not the other person. And that's an interesting thing because what we found at Airbnb was that the quid pro pro works a lot better saying, oh, you get $5 and your friend gets $5 as well, worked a lot better in terms of messaging as well as the strategy. So for Robinhood, what we can assume here is that the users are pretty selfish and this is fine. Like I'm not, being selfish isn't a bad thing, but this is what they found. So if you go get the link and you share it, the first thing you see that, like let's say I sent to my friend Holly, Holly goes and looks at it and it says, your free stock is waiting for you. Again, they're playing back into the selfish bit. It was like, oh, you need to go get this free stock. And then the other thing is that it says, Brian invited you to join 18 million plus people Robinhood. This is just social proof, sometimes it works, sometimes it doesn't. This is pretty nice. I do want to point out their CTA, which I know is a micro optimization, but it's just fun, right? It says claim your free stock, which is very, very different to a user than get your free stock because claim means that you already own it, you just need to grab it where get it feels like it's a little bit more of an activation energy. So they've definitely optimized the hell out of this page and it's very interesting to see what they've done. So then it says join your way, but first your email, they ask you to input your email and you hit continue. And then when you hit continue, it takes you to just the sign of page. It doesn't do anything. Like if you look up, they ask me my first name, my last name and then my email address again. But by getting your email address here, what they get is now they can go and email you saying, hey, you forgot your referral. Remember to go claim your free stock because now they have your email address. They know that you came into this page and now they can't get you back up. So I would say that for the most part, Robin Hood's referral strategy is pretty solid. The only thing that I would really worry about is how they go and look at their email address because if you ask for the email address back here, just fill it back in once you hit the sign up page. And then I ran one referral strategy at Haven and Haven was a fintech company. You can go and store a bunch of money in the bank. And we did it so that every single time you refer to a friend, you would get an extra 0.25% of APY up to 4%. If you don't work in fintech, a 4% APY is ridiculous. That's a lot of money. But this didn't work and it didn't work because there was absolutely zero dopamine hit for getting 0.25%. What really works is do not delay the dopamine hit. People want money or a monetary equivalent. If you can give them money, if you can give them $10 for signing up, give them $10. If you can give them $15 in Uber credit, give them $15 in Uber credit. But really the monetary reward is what drives referrals. And if you do not give people money, it's just not going to work. Be thoughtful in your reward system. Again, I mentioned the selfish, altruistic, and quid pro quo. Think about how you want to position yourself in your brand when you do that. And referrals is really, really, really good early on. But I would say as you get bigger and you hit market saturation, you do need to revisit referrals in order to figure out the incrementality of this. And then the last thing I want to talk about is activation and engagement. In specifics, I'm going to focus on both email and push. So for push and email, there's different levels of urgencies based on the different channels. So for push, push is annoying. It's like, oh, this is a problem that you need to go look at immediately. Whereas email is a little bit more delayed, so you can go and take your time with it. So be more selective when you send out push notifications because you can push for a lot of user interactions, but you can also get them to unsubscribe if they find it too annoying. And the other thing that you should definitely be thinking about is you should be thinking of everything in terms of user life cycles, right? So everybody knows if you're going to send an email, send it to specific messaging to people that are 18 to 25 and 26 to 32. But it's much less about thinking about that than more like, where are your users in the life cycle of your product? A user that signed up a week ago is expecting very, very different things than users that signed up six months ago. So if you haven't defined user life cycles, I would 100% do it because not only is it helpful for other areas of growth, it's going to be very helpful as you go figure out this push and email strategy as well. And then last two things I want to talk about, unsubscribers are forever. I don't think, if a user unsubscribes, I don't think there is a way to get them to read, subscribe back to an email or push on a vacation. People just, people don't like emails. So whenever you burn a user, they are burned forever. So be very, very careful in how you're doing this and if you're annoying the hell out of your users because you don't want to do that. And then I think everybody knows this, GDPR is scary as hell. Be very careful how you're storing your user data and how you're sending out emails and make sure you don't break into the laws because it will get very, very, very expensive. So the biggest tip that I have that I got from my friend was that you need to send users the right content at the right time. And this is really important, especially when you think about low and 10 emails. So low and 10 emails are emails where you send users an email and hope they do something when they didn't ask for it. And this didn't work at Airbnb because like what a low and 10 email looks like, it says take the trip less traveled and they gave me a bunch of suggestions. One, I didn't ask for a trip. I'm not planning for a trip and me going for a trip is not related to Airbnb at all because traveling is rare. Traveling doesn't happen all the time. And even if we show nice listings, the user quality is gonna be super low and the conversion is gonna be super low. So what you end up doing is you're just spamming your users with like, hey, look, check out this cool thing that we have and they're gonna be like, well, I don't really care about this and they're gonna leave. They don't travel because they see nice listings. They travel because they want to. So if you look at this email, it's just not a great email to be getting. Instead, what I like to focus my time on is on event-based emails. This is where you see the most success because this is where users have engaged and shown intent. So like if somebody is on your platform and they're interacting and they're doing something, like send them emails based on that information. So I think like the most recent one I got that was useful is this eBay one because I was looking at climbing shoes and I forgot about it and just left it alone and then I went and continued to do something else. But eBay knows that I looked at this climbing shoe. They know that I'm interested in this on some sort of level and now that there's only one left, they send me an email saying, hey, come back to it. Why don't you go look at this shoe again? And like this is actually helpful because I have shown interest in this. I want to know when this shoe is going to be available and now they've told me that, oh, there's only one left. So these event-based emails are really, really critical in order to engage with the users because you're providing a lot more information to them than you would with like these really terrible low-income emails. Nobody cares about this. People care much more about this. So lessons, push notifications I've learned should be pretty consistent sends. Like if you were able to send consistent, valuable push notifications, then people will find value out of it and they won't unsubscribe from the performance-intentful action. The best time to engage with the users is right after they sign up because they've shown a lot of intent. They know exactly what they want from your product. They've done something, now you can start walking them through the process of what you need to do in order to get the most value out of your product. Taiwan Platform matters for users whenever you engage with them, right? This goes back to the user life-cycle stuff that I was talking about and you should send things that matter. And the next slide is actually my favorite slide. I interviewed my friend that worked, was a PM at Airbnb on emails. So the first tip she gave me, don't send fluff garbage, which is pretty obvious. But then everybody likes to go off my subject lines. She's like, subject lines don't need to be short and don't waste a ton of your time on this, test like three. And my favorite line ever is the best time to send an email is when you have something fucking relevant to send them. So if the most relevant email you can get, you're sending is Friday at 5 p.m., send that email. So in conclusion, grocery lay done as foundation, it can always be optimized later. And this is a really important thing to be thinking about is that you can go and spin up your SEO thing right now and let it sit around for a couple of weeks or a couple of months or even a year or two and that's fine. Know what a balance opportunity and optimization because this is really important when you're a big company and you're a small company. If you're in an early stage startup and you spend too much time, you know, like optimizing running A, B tests, you're gonna get lost in the weeds. And I've seen tons of companies get really lost because they wanna go and test everything. As a small startup, the point of your product, what you wanna be doing is you wanna launch things as quickly as possible, right? Like the example I like to use is if you go checkout, if you run an e-commerce website, you're not gonna go A, B test a checkout button. You need to have a checkout button. So if you're a small company, don't get too lost in the optimization, running A, B tests, launch things and launch them quickly. And yeah, the last thing that I wanna go over is my favorite thing about growth is that you should ship fast and learn quickly. You should always be iterating. And if you ever fail, that's fine because failure is how you learn. The biggest problem I find with growth teams is once people are scared to fail, that's when your company starts to stagnate. So as long as you go and are always trying to learn from everything that you launch and every experiment that you run, you're in pretty good shape. And that's it. Thank you for your time. Hopefully that was pretty helpful.