Casey Winters

Growth Expert and Advisor
Greylock Partners and Reforge

B2B Growth Loops: A Reforge Masterclass

Casey Winters: So a little bit about me. I basically work on helping companies scale. So as an employee, I led the growth team at Pinterest. I started the marketing team at GrubHub as a growth advisor in residence at Greylock. Uh, and I started my career as an analyst at and now I work full time as an advisor to startups and growing tech companies on how to scale their businesses. So Eventbrite, Thumbtack, Rented - a bunch of others. So I'll start with a simple question. How does your product grow? And it may sound like a simple question, but my guess is if you go around and ask different people at your company, you're going to get some different answers. Your product team might say, oh well we build a great product and everything else sticks. It takes care of itself. Your finance team, I talk about how you make money or your business model and your marketing team might talk about some sort of funnel like Dave McClure's AARRR.

Casey Winters: And what I find with companies that I talked to you when they've actually asked this question amongst themselves enough, they usually center around some sort of funnel. So you have people you acquire at the top of the funnel. You then do work to activate them onto the product value. You then make sure they retain, make sure they make you money and then make sure they tell other people so that you can get more people at the top of the funnel. And um, this is a useful framework but it's about 11 years old and we find that it's not quite representative of how the best companies actually grow and that there's some real problems with it and implementation. So one problem is the parts of the funnel tend to be owned by different parts of the company. So you might have marketing in charge of acquisition, you might have product in charge of a retention, you might have sales in charge of revenue and they prioritize different metrics and instead of working together, they actually optimize a bit against each other where they...where they complain to each other that each one is the cause of the other people not hitting their goals.

Casey Winters: Oh, why didn't I hit my sales goals? It's cause I got shitty leads. Or Oh, why are we not making off revenue? It's because like the product's not good enough, et cetera. So not only is this structured, where teams are competing, are working against each other. Um, it also inevitably means that the problem eventually becomes, I need more people at the top of the funnel. I need more tactics to get new users. I need more channels. I need more money to spend more and more and more. And that's because this funnel thins out at the bottom so you don't get enough referrals to kind of replace the people at the top. And would almost every company that I ended up talking to to potentially advise this is one of the top problems they bring up is more people at the top of the funnel. And I'd argue it's, it's basically the wrong question to be asking.

Casey Winters: Funnels are really good at driving linear growth, but the best companies in the world don't grow linearly. They compound over time, they have this exponential growth and they don't get it by optimizing a funnel. So here's kind of an example from my career. So this is a stock chart of Grubhub in blue and BlueApron in red. And you know, at this point, Grubhub had been public for a few years and blue apron went public in 2017 and you would think Blue Apron lot younger company, a lot more growth ahead of it. It should be growing at stock a lot faster than grub hub. But over this period of time, Grubhub's stock grew 85% and blue apron stock dropped 70% in. Why is that? It's because Grubhub was built to compound and blue apron is optimizing a funnel to grow linearly and eventually they ran out of top of funnel channels to plug in profitably to grow their business.

Casey Winters: Their growth rate stalled, their stock price drops dramatically as a result. Whereas, you know, this is maybe year 14 into Grubhub and it's still growing quite fast because it's got ways for it to compound. So if funnels don't work to represent how your company actually growing, what's a better way to do that representation? And obviously based on the title of my talk, my answer is what we call a loop, where you have some sort of input generating an action that creates output that can reinvest into the input over time. So a technical definition of this is a growth loop is a closed system where the inputs through some sort of process generate an output that either directly reinvests into creating more input or is more input in and of itself. And I get that that's a little bit technical. So we're just going to go through a lot of examples to make it clear.

Casey Winters: So this is an example for my career. You know B to c example, but don't fault me for it, which is, this is one of the main things I worked on at Pinterest when I got there. So in this case we have people see content on search engines. They like that content and they sign up for Pinterest to see more of it. Then when they sign up for Pinterest, we get them to see a lot more interesting content and ask them to save it for later and we ask them to get an extension to bring in new content that they like into Pinterest. A bunch of these users end up creating boards by saving new content into Pinterest or saving stuff that they see on Pinterest, which we call a re-pin. That creates new boards, new content that we then distribute to Google, which then gets more users.

Casey Winters: So essentially our retention, which is getting people to bring in new content or to save content that we show them, creates its own acquisition opportunities by distributing that content to Google to attract new users who are typing in, you know, chuck a boots or wedding ideas. So in this case, Pinterest grows. It's new users by focusing on getting more people to save. And by doing that, we then distribute that content to Google, which takes care of acquisition. So then if we just make sure we're really good at this, every other part of the loop works to compound the business over time. And this has been the strategy we've used to grow Pinterest from let's say 40 million monthly active users to now well over 250 million monthly active users, no money on performance marketing, no sales, none of that stuff. Um, just what we call a user generated content loop.

Casey Winters: Now that is one type of a bunch of different types of loops that I'll go over today, what we call a content loop, but there's two other main types then we're gonna talk about. And then some kind of subsections. So the first is what we call a viral loop. That's probably what most people are familiar with when you think about loops in a viral loop is constrained and also fueled by how many users are in a system. Because in a viral loop, your users tell other users to either use the product more or to sign up for the product more. So if you don't have a lot of users, you're not going to compound a lot. If you have a lot of users that tell a lot of other users, then you're going to grow up really fast. A content loop is not actually constrained by the number of users in the system.

Casey Winters: Though a content loop is constrained by how much content there is in the system. That's what's going to fuel the growth of the business. So at Pinterest, there's now over a hundred billion pins in the Pinterest ecosystem, all of which are doing work to attract new users who are going to pin more things in a paid loop. The constraint and the fuel of that is capital. I need to be making enough money that I can reinvest in different initiatives that are going to drive, you know, more new users. So let's, let's get to kind of some examples. First, I'll start with kind of viral loops. So viral loops, there's three different designations that I want to talk about. The first is a personal viral loop where users are inviting other users to the product because they feel like the product's going to get better for them. So I invite my coworkers a slack because then we can communicate about work on slack.

Casey Winters: A financial viral loop is where I'm inviting another user because I get a financial benefit. I'm going to get a discount on this product. And a social viral loop is I'm telling other users about it because it's going to make me look cool usually because the product quality is so good. So what's an example of a personal viral loop in B2B Front's a recent example that I find really fascinating. How many people are familiar with front as a product. Okay. A lot of you. Um, so front is basically a shared inbox for customer service teams. Sales teams, et Cetera. So the whole goal of the product is you sign up, you use it to communicate with your customers, and then because you're not going to be the only person that communicated with that customer, you end up, you end up inviting your coworkers on your support team or your sales team to say, hey, here's the communication history I've had with this customer.

Casey Winters: If they write in and I'm not around, you can pick up exactly where I left off. So the whole value of the product is inviting coworkers to the product who then are going to use the product and then they're going to invite more. So front grows virally within organizations very quickly. And then as you know, support teams, sales teams, switched jobs, um, they then help it grow virally outside of the organization, right? Where if I used front at my last company in and I take a new job, I'm more likely to bring front into that organization.

Casey Winters: Gusto is a good recent example of a financial viral loop in B2B. So you know, a new user, typically a small medium sized business, they sign up, they use Gusto to manage their payroll software, and then Gusto asked them to refer any other business owners they know to get a discount on how much they pay Gusto. And because every business owner is financially motivated, they take people up on this and they usually talk to a lot of other business owners. So they have the means of reaching other people. Now there are some issues with this and B2B in that for Gusto at works because owners are actually motivated by money, but employees that companies aren't, because if they're going to save the company $100 on software, that $100 doesn't go into their bank account, it stays, uh, in the company's bank account. So usually a financial viral loop in B2B is more useful when it's someone who feels that ownership responsibility that is going to refer if, if it's not an employee, they need to think about is it gonna be clear how the product gets better for them? Or is it going to be clear how them recommending the product is going to make them look cool?

Casey Winters: So stripes of an example of what we call like the social viral loop or word of mouth loop, where the user likes the product so much that they're incentivized to tell others because then they're going to get, they're going to accumulate some social capital in the process. If you were a developer that told people about integrating stripe versus using or PayPal or something, and all of a sudden your process of integrating payments goes from like months of hard development work to like a couple of days, uh, then you're gonna look awesome in your peer community. So stripe grew in the early days very quickly via word of mouth, uh, and, and this loop is sustainable. There's two types of word of mouth that I think it's important in a B to B business to understand where you are. Is this the type of product where it was so great that people are just gonna want to tell a ton of other people about it?

Casey Winters: Or is this the type of product where naturally people aren't going to tell a bunch of other people about it, but when other people have this issue, they're going to solicit opinions from people they trust. So in poor word of mouth, it's like you're buying a home for example. It's a very big decision. You're going to ask everyone around you like, Oh, where did you find your home? Did you use a realtor? Et cetera. And they'd be like, no, I used red thin. It's awesome. Here's why. Um, so think about if you're going to try to grow via word of mouth and B2B is going to be the thing where the product experience is so great that people are just going to use it and then tell a bunch of people. Or is it going to be when people have the need, they're going to solicit others and need to make sure those people have experience with your product and prepare to communicate its value.

Casey Winters: Everyone wants to grow via word of mouth. It's really hard. And one of the reasons it's really hard is, uh, first off, you need to have a 10x better solution to grow via word of mouth. You need to essentially be remarkable, which means there's some expectation or alternative on the market where people are solving their problem and then you deliver something so much better that people are just so amazed that it's even possible. And this is great. This is what fuels word of mouth for most products. But what happens is that over time that expectation increases new people come into the market that copy what you've done and then all of a sudden you're not 10x better than them. Or because it's become so aware that you are the standard. It's no longer remarkable that you're so much better than everything else because everyone uses you already. So it's very hard to keep these word of mouth leaps going because you need to consistently stay 10x better than the market and staying 10 x better than the market is very difficult.

Casey Winters: I love this quote from Jeff Bezos with illustrates this, which is, you know, he says, one thing I love about customers is that they're divinely discontent. Their expectations are never static. They just go up. It's human nature. We didn't ascend from hunter gatherer days by being satisfied. People have a voracious appetite for a better way. And yesterday's wow quickly becomes today's ordinary. So this is usually the issue with a word of mouth loop is that yesterday's wow quickly becomes today's ordinary. So we find a reforge is that the personal viral loops where the product gets better, if you invite others or the financial viral loop where you have some incentive to share, tend to work better in the longterm than word of mouth. And of course, any word of mouth you can get you all take, but it's hard to rely on that longterm to grow your business.

Casey Winters: Okay, so this is the different types of viral loops. What about content loops? So if viral loops are differentiated by who's who's distributing and what their reason is for it, content loops are more differentiated by who's actually generating the value, who's generating this content? And there's the company, you can create the content, which we typically think of as content marketing users can create the content like in the Pinterest example, or if you're a marketplace, the supply side of your business potentially can create the content. So let's look at some examples. Hubspot's probably one of the most famous content marketing loops. It's fairly simple thing to understand. They create content, they get an index by Google. People visit the content, some of them become leads, some of them end up making the money come in, making the company money, which then they reinvest in more writers, creating more content.

Casey Winters: Now the question is like, okay, but like hiring writers is really hard. They're expensive. And then they run out of interesting things to write. How do you actually scale content marketing for the longterm to grow your business? And usually the answer is you start hacking the distribution of the content and the creation of the content. So Hubspot got really good at not just getting Google to index pages to drive sharing, but they also got users to start sharing it. Marketers are motivated to be seen as people that understand their field, that know all the latest things. So they're incentivized by social capital to distribute this on Twitter, on Facebook, on LinkedIn, et cetera. So this is a way where they're getting the loop to turn faster with the same amount of content. But the other thing that Hubspot got really good at is also getting other people to write content. So then instead of the company being in charge of writing all the content, they're actually having guest posts from the community, which you know, we consider supply and now they have this augmented inventory of content that gets distributed to Google and get shared by users. So usually for content marketing to scale, you need to have multiple points of distribution and you need to have multiple points of creation.

Casey Winters: But that's digital content that's actually like writing something. One of the things that I found fascinating about, um, content is that it can be physical as well. So square I think has been a good example of using what I'd call a physical content loop in that a business joint square, they deploy their point of sale system. Thousands of people interact with that point of sale system. A small percentage, which are business owners who use that experience and like, wow, this is great. Why am I not using square? And then they end up signing up in an order for a physical content loop to work. The physical content needs to be super noticeable, right? So like squares put a lot of effort into their physical product design so that when you come up to this and you use it and you're like, wow, that's amazing, that's beautiful.

Casey Winters: Like if I'm a business owner, like why am I like using technology that looks like it was invented in the 80s when I could be using this. So you know, we see this in B to c businesses with things like scooters, but it's also super powerful and B2B businesses where you use this thing and if you're a business owner, you're like, wait, why am I not doing this user generated content? Something that's a little bit newer and B2B, but something that's been fascinating. How many of you are familiar with g two crowd? Okay. Wow. Okay. They're doing really well. Um, so G2 Crowd's pretty awesome in what they do and that they have users write reviews on B2B software, um, email marketing solutions, payment solutions, et cetera. People have used those products, right? Their experience experience to help other people decide if that product's right for them.

Casey Winters: So when a review gets written, they distribute a page about that review to Google people who are searching for like, should I use exact target or MailChimp for my email marketing? Find that content on Google. You know, they sign up and they usually have expertise on other things. They might be looking for expertise on email marketing, but they have, you know, expertise on lead software or something else. So then they distribute, they write a review, which then helps g two crowd get more content, distribute more content to Google and acquire more users. So, even though it's a B2B example, very similar process to Pinterest in terms of how G2 Crowd goes DocuSign's a slightly different example and user generated content and that news or sign up to create documents, the whole purpose of creating that document is to distribute it, to get someone else to sign it. So in the process of using DocuSign, you distribute it to someone else who then signs up and then they've used the product and then they're like, wait. So I could be like asking people to fax and scan things all the time or using this, I'm going to use DocuSign the next time that I need to send a contract. So this has been a fascinating way for DocuSign to basically grow their awareness, grow virally. I'm in a space that's typically been very, very sales driven.

Casey Winters: Um, switching to supply generated content. Eventbrite's been a company that I've been working with a lot recently and I find that their loop is very interesting in that, you know, event bright acquires, event creators who then go and create events on the platform to accept tickets. So what happens is once you acquire a new event organizer, their whole thing, their all reason for being is to create an event. But what are event organizers called? They're called promoters. They know how to reach their audience, right? This is what they do when event vent is providing is a way to easily accept payments. But the event creator already feels responsible for their own distribution of an event. So they start distributing their event to their fans that they already have to potential fans that they can acquire, which then creates a lot of attendees from a typical event like this.

Casey Winters: And then only one of those attendees needs to say, wow, like purchasing on event. It was really great. The next time I have an event, I'm going to use Eventbrite. And then they become an event organizer. So this supply generated content loop has been the primary path of growth for Eventbrite from a startup to now a public company. And it's, you know, entirely free event creators drive more event creators. Um, because it turns out event creators also go to a lot of events they go to. In fact, they go to more events than people who are not event creators.

Casey Winters: One of the things to think about for a supply generated content loop is ease there. That motivation of the supplier to actually do their own distribution. So for Eventbrite it was very clear, right? Like I create an event and then my goal is to get as many people as possible to it. It's this one time thing, but Grubhub's value prop is that I helped delivery restaurants get more delivery orders. So there's no incentive for a grub hub restaurant that they sign up to then tell all of their existing patrons to order from Grubhub. All of a sudden you have, you know, these people who are getting orders for free and now they have to pay Grubhub 10% right? So there's no value promise to, you know, a restaurant to say like, Oh, now that I'm on Grubhub, I'm going to tell all the people that already dined at my restaurant to use it. The whole point of me using Grubhub's to get more people who don't already dine at my restaurant. So in the case of grub hub, like their distribution needs to happen through Google. It needs to be owned by Grub hub, right? Whereas for eventbrite's, we basically have decentralized that saying, hey, event promoter, you're good at promoting your events. Go do that, will make accepting payments really easy. And obviously you'd prefer that, like your supplier is helping you grow versus you have to do all the growth yourself.

Casey Winters: Survey monkey is a very similar example actually to Eventbrite, where they have people come in and create surveys and then they're distributing these surveys largely to thousands of people. Only one of those people who sees that survey has to say like, oh, I sometimes need survey help. I'm gonna use survey monkey the next time I have a survey to have a sustainable compounding loop over time. And Similarly, survey monkey or recently public company, the primary way that they've grown and gotten distribution is through the surveys themselves. This is something that they don't have to spend money on. You don't have to do sales, et cetera. People find out about it naturally from receiving surveys.

Casey Winters: Drift and Intercom are more recent examples that I found interesting in that they're actually created, they're actually embedding software onto sites to help people become aware of the value of their software. So in this case, a new user signs up for drift. They create a chat bot on their site, a bunch of people, they embed the chat Bot on their site and an a bunch of people end up interacting with that chat Bot and being like, wow, this is a really great experience. What the Hell is powering this? They click on the drift branding, they click on the inner calm branding and then they end up getting drift or inner calm themselves. So thinking about like, is there software that's embedded that will become distribution for you? It can be something that's very powerful.

Casey Winters: Lastly, talking to paid loops. A lot of you are going to be familiar with these, but I'll go through them very quickly. Um, paid loops are differentiated, but based on how much friction is there to understand the value of this product and that will therefore tell me who needs to distribute the value. So in the case of an ad loop, determining the value is not very difficult for the user. So you just write an ad, it says what it does and people are like, oh I want that. And they click on it for a sales loop. The value promise usually is something that takes a lot longer to understand and it might need to be customized for each individual customer. So therefore you typically hire someone at your company that's in charge of doing that for each person that they talk to. And an integration loop is where the value promise is the connection of your value with another company.

Casey Winters: Therefore, the distribution of that has to be through the other company because there's something that you do together that's unique. So let's walk through some quick examples. So if in Zapier, you know, example of they can use paid acquisition to grow, they can acquire new users, they can create revenue from them and once they make enough profit from that revenue, they can use some of that revenue to buy more ads. Fairly simple. Most of you are familiar with this, but all I want to call out here is that in order for this to work, uh, you need to be making enough profit from your lifetime value to reinvest more money in advertising or else you will eventually run out of money. And what typically happens over time is the first people you target with paid acquisition, they have their exactly the, the right fit for your product, right?

Casey Winters: They're the right target market. They have a high conversion rate, which therefore means a low CPA and then they have a high lifetime value. But over time as you need to expand into finding more customers, you're then moving drifting further and further away from your target market. You're getting lower conversion rates on your ads, you're getting therefore higher CPAs on your ads and you're getting lower lifetime value on your ads because they're not as good a fit for your product. And then eventually that gets upside down where your CAC is less than your LTV. So I don't love paid acquisition as a means to grow for B2B or B to c businesses in unless you have some other thing underneath that's working really well in paid acquisition is just fuel to help you grow faster. If you have a data network effect for example, or something like that, then I throw like as much, um, performance marketing as I can to fuel into that.

Casey Winters: If I have a cross side network effect because I'm a marketplace, I might use that. But otherwise paid acquisitions, just a good accelerant to something else. It's not usually something super sustainable on its own sales. However, it can be very sustainable. So most of you are familiar with sales, but I'll just cover it quickly. You know, you have a sales person's ad for Grubhub, for example, sign up a new restaurant. That restaurant starts making you a bunch of money through transaction fees. And then once you've made enough money and transaction fees, you can hire additional sales reps. Super sustainable. Um, sales reps are not cheap, but as long as you have a high enough lifetime value from the people you're signing up, then you can afford, um, to scale up a sales team and hire more of them.

Speaker 3: Uh,

Casey Winters: but you can also kind of outsource that sales effort, right? So zero is an example. If they don't hire a salesforce directly, they actually work with value added resellers and they start scaling that more over time. Completely sustainable as well. You know, Jira is another example of this where they don't have a direct sales team but they use a lot of value added resellers. Totally fun. Shopify is a good example of an integration loop and these are something we're starting to see a lot more with the latest crop of B2B companies. You know, segment's been a good example of this as well, where they are creating an integration and then when the integration is created, both Shopify and the person and the company integrates with, let's say stripe are now both incentivized to promote this integration. So all of a sudden you now have a partner that is marketing Shopify for you.

Casey Winters: When they do that, they sign up a lot more new customers do they make a lot more money from, which means then they can do more business development to create more integrations. Uh, so Shopify has done a good job of this segment's done a good job of this and we're starting to see this become a more common strategy to grow. Um, and it's something that you should think about. Like are there other companies where one plus one equals three and will those other companies, if that integration is created, being incentivized to talk more about us in a way that's going to drive more usage of our product. So these were just a sample of a bunch of the different loops that we've covered in reforge on how to grow B2B businesses sustainably, how to compound them over time. If you want to learn more, you actually teach an eight week course on this, um, called the growth models course at reforged. So would love to have you on the course if you're interested. And with that, I'm out of time. So I think we have a little bit of time for questions.

Speaker 1: [inaudible] hi. You mentioned Slack earlier, how the loop was more about growing within the company and its early days, you know, without it getting tons of funding. Was it [inaudible]? Was it primarily word of mouth would you say or?

Casey Winters: Okay. Yeah. So, so one of the things that we talk about in the chorus is that these different loops have different minimum and maximum scopes and that a lot of these loops won't work day zero when you launch your company, right? They need to have more users, I need to have more content or et Cetera, or you need to have more money in the bank. Uh, so in the case of slack, slacker early on via word of mouth and then once inside a company it grew virally. And then of course in silicon valley, engineers, designers are leaving companies every day and they tend to bring slack into new companies with them. Uh, and now when you look at slack, slack is more a of what we call, like it has a full growth model of a bunch of different lutes working together. It has sales, it has paid marketing, it has by reality and it, and it has word of mouth and all of these together kind of create a compounding growth rate for the business. But early on I it was largely driven via word of mouth.

Speaker 1: Yeah. I totally agree with your presentation.

Casey Winters: Thank you.

Speaker 1: Um, I was wondering when you hear from investors what your plan is for paid customer acquisition? Yeah. Like what are some, I mean aside from slack, which is like the most obvious and it's really easy to explain, like, have you, have you, do you have some data points around how investing in those other areas of the loops that what would be more effective? Cause I, I still talk to a lot of investors who expect you to grow via paid. Yeah. And I think it's totally wrong and it's very difficult conversation to have with an investor.

Casey Winters: Totally. So, so A, there's a lot of variability in the quality of investors. So, uh, and it also depends on what stage you're, you're investing in. So, you know, at gray lock, we primarily work on series A and B. And for those stages, every investor is primarily trying to answer three questions, which is A, do you have strong retention? B, are you going to be able to monetize that retention? Well, you don't necessarily need to be doing it today, but I need to believe that at some point you can monetize that retention well, and then three, does a and B create a sustainable acquisition strategy in some way? So if you are, um, retaining people really well and you're monetizing them really well, then you can just talk about reinvesting those profits into paid. But let's say you're not paying, you're not getting, um, you're not getting Ma, you're not monetizing yet, which is how Pinterest was when I joined. They needed to have a credible, um, strategy that you're either proving or that they believe in as investors on how you're going to be able to grow without pain. So in the case of Pinterest, you know, initially we grew very virally and then once the Facebook open graph turned off, like our viral growth essentially stopped.

Casey Winters: And that's when I started working at the company and started working on creating this user generated content loop via Google. Um, so what would investors really want to know is what's the scale of black position strategy and they expect it to be paid. If you had a, if you have a good example that you can model and show how it's working without paid, they typically don't necessarily need paid. Um, but even if they're pushing on paid, then you just want to show how paid can be an accelerant to this thing that's already working. So in the case of Grubhub, where we have a user journey, we have a supply generated content loop was the primary way we grew, but then that creates a cross side network effect where once restaurant sign on, you know, people are, have a higher conversion rate. And then once people have a higher conversion rate, it's easy to sign more restaurants. Um, so essentially when we're raising money, we're talking about, hey, we use paid to accelerate this network effect. But what's important is the network effect. I'm not, not the paid itself. Does that answer your question? Okay.

Casey Winters: Oh, he's got notes.

Speaker 1: Yeah. Got Notes. No, I love the presentation. That's loud. How would you relate the micro loops to Nir Eyal's? The Hook canvas if you're familiar.

Casey Winters: Yeah, I would say. Um, so for those of you who aren't familiar Nir Eyal wrote a book a few years ago called Hooked. Um, and I'd say Hooked was an early example of the loop model. Um, and the hook, this specifically, um, the loop model when it comes to retention loops and engagement loops switches, once a person's using your product, how does that usage essentially create more usage? And there's different ways that it can do that in what Nir focused on is essentially how people, when they use a product, there's some sort of user to user based incentive to bring them back that locks them into the platform. So if I write, if I post something on Instagram and then people like it, I'm incentivized to come back.

Casey Winters: Um, that is one of many ways that usage can drive more usage. So at Pinterest for example, that's how it worked in the early days. And there's a bunch of Pinterest examples in the hooked book, but we actually scaled out of that working because what we found over time is that what was motivating people to use Pinterest was less and less what your friends were doing over time or less and less how much engagement you were getting from the community and more and more about your personal interest. So people actually stopped caring if other people were liking their pins or repinning their pins and more about, am I finding cool content on this platform? So Pinterest now drives retention via a data network effect, not via something that looks more like the hooked model, um, where the, the more stuff you save with Pinterest, the better Pinterest gets at recommending more content that you're likely to enjoy.

Casey Winters: Meaning, your chances of repenting go up, your chances of coming back more often go up and didn't. That work affects are super powerful and B2B as well. Um, you know, for example, every additional transaction that goes through stripe means they get better at fraud detection for every single one of their customers. Right. And that's something that like, even if you don't like stripe for other reasons, you might not lead because then all of a sudden your fraud is going to jump like 25%. Um, yeah, so I'd say like the hope model is just an example of a type of loop that described a lot of how the social companies were growing over the last five years. So I'd say it fits very nicely into the, the growth model, um, of course that we built, but it's one specific subset,

Speaker 4: uh, on this slide that you use right before Jeff Bezos. Amazon, you mentioned like the, the new tool that you create has to be 10 x better. Yeah. But what if you're solving a problem? I think you used Zapier as an example, like to get 10 people hate creating custom integration. They use Zapier 10 x better, but what if then Zapier itself gets annoying because it's not bilateral. So then you end up having to create custom integrations anyways. It's like, what if, so what if when your 10 x becomes your base, like what would the next be?

Casey Winters: Yeah. Um, so usually at this point that's when like word of mouth growth starts to drop. Um, essentially finding... Sachin just gave me a hand signal, but like I only caught like a second of it, so I'm not sure what he was trying to get at. Um, maybe he's telling me to wrap up. Uh, but uh, essentially we were trying to find is like what's that next product innovation that's going to keep my 10 x? Like how do I go for Uber x Uber Pool as an example. Right. Um, and if that's not going to be the thing, then what is this word of mouth bootstrapping some other loop that now I've hit the minimum scope for. Now that I've got a bunch of word of mouth, I have a bunch of content I can distribute or now that I have, I've got a bunch of word of mouth, I can build a viral loop on top or I can make a bunch of money and start hiring salespeople or performance marketers.

Casey Winters: Um, so that's generically how I'd think about it. But I think embedded in your question was something a little bit about like these integration loops, how to actually get them going. Um, and usually the way companies do that is they build all the initial integrations themselves without really talking with the other companies. And then it becomes more of an integration loop model over time where you start outsourcing, building the integrations to the company that wants to integrate. So segment initially just as all the integrations with, you know, the web analytics companies, whatever. And then once they get big enough people start coming to them and saying like, I want to integrate with you because it helps me sell my product into a other companies. And then segment can start to like decentralize that and be like, yeah, I use this set of API as knock yourself out. Well, we'll send an email to our users when you sign up. Stuff like that. Okay. Am I done? I'm done. Thanks so much for your time.

Speaker 1: [inaudible].

Keep me posted on Empower 2019.