The SEO Framework That Propelled Pinterest To An Estimated 72M Users with Casey Winters

So, you know, it definitely becomes a race of innovation, and you can't say, well, because we're focused on this and this other company's focused on that, we're going to win. Because if you start winning, then everyone else is just going to respond by adopting your tactics.

Do you want to impact the world and still turn a profit? Then you're in the right place. Welcome to Growth Everywhere. This is the show where you'll find real conversations with real entrepreneurs. They'll share everything from their biggest struggle to the exact strategies they use on a daily basis.

So if you're ready for a Hey everyone, just a quick heads up that we're giving away an ebook called 29 Growth Hacking Quick Wins. We co-authored this book with Matan Griffel of One Month, and it'll give you a solid base on where you can create growth ideas from. So all you need to do is text QUICKTIPS to 33444. That's the word quick, Q-U-I-C-K, and tips, T-I-P-S as in sugar, to 33444, and you get instant access. All right everyone, today we have an awesome guest. His name is Casey Winters, and he leads growth over at Pinterest, which is a visual bookmarking tool which is on fire. Unfortunately, they can't reveal or disclose public numbers, so I'm just going to give some estimated numbers right here around their user base, which is estimated over 72 million users. And they have over $1.3 billion in funding. And the numbers that Casey could share, he graciously shared with me, over 50 billion pins and over 1 billion boards. He's also the first marketer, or was the first marketer at Grubhub. Casey, how are you doing today?

Doing great. Thanks for having me.

Yeah, thanks for being here. So why don't you talk to us a little bit about your, who you are and kind of your journey through Leading growth at all these awesome companies.

Yeah. So I've been working on growth-related things for a little over a decade now. Started my career working on marketing analytics at a company called Classified Ventures. They did basically online classifieds, so apartments.com, cars.com. It was just a great way to learn about what are the things that actually grow internet businesses. Naturally, the next thing I did is after I learned how to measure, those types of initiatives. I started working on those types of initiatives. So paid search, organic search, email marketing, those types of things. The cool thing about Classified Ventures is that they were starting up a lot of new companies inside. They were doing a last-minute apartment startup, a rental home startup, and then eventually a real estate startup which I went to lead online marketing for. It was really a great building ground to just learn how to grow a bunch of different types of internet businesses. After that, I started at Grubhub as the first marketing person. They raised $1 million in Series A funding. We were in a couple of cities and they really just needed someone to figure out how to take this thing national. So I spent a lot of time just figuring out like how do you grow a marketplace that has restaurants and also people that order from those restaurants. Spent a lot of time focused on growing the consumer side of that business. So built out organic search strategies, paid search strategies, email marketing, conversion optimization, loyalty programs, transit advertising, TV. You name it, I probably spent some time on it or spent some money on it. Really great experience. We grew from 3 cities when I started with the 3rd city getting an order a day to over 600 cities, went public in April of 2014, acquired a bunch of companies while I was there, and now is the dominant national player in online food delivery. They're valued at over $2 billion today on the public markets. Decided that I was a little bit through with the cold of Chicago, which is where Grubhub was. Started looking to sunny California at opportunities and landed at Pinterest working on growth here. Now I lead the growth team, which is a cross-functional team of engineers, designers, product managers, and analysts. Really, we're focused on international growth these days. Pinterest is 40% international service. We're really looking to— make it a majority international service. So spending a lot of time on figuring out the right ways to educate, figuring out how to get in front of people across the world and explain what Pinterest is. And yeah, it's a ton of fun. I'm really enjoying it.

Awesome. Now, your experiences, I mean, having to do classifieds and doing Grubhub and then Pinterest, I mean, it involves a ton and ton of pages. A lot of different listings. So, I mean, what are some trends you've seen growth-wise that have kind of carried over from startup to startup? I'd imagine SEO is a big player. Why don't you just talk a little bit about that?

Yeah. So I think you correctly figured out that when you have a lot of content and a lot of pages of content, search can be a pretty strong way to try to grow a business. So in the case of Classified Ventures, we had All of these listings and all of these neighborhoods that we had apartments for or homes for. So the kind of building block of building an organic search strategy is how do I organize this content in a way that people are searching for where if they landed on my page it really helps their experience. So in the case of Apartments.com, someone searching for Los Angeles apartments They land on a page that has Los Angeles apartments, that has local information about Los Angeles. Figuring out how to do that without hand-curating every single one of those pages was a big part of what the team worked on. It's a very similar process when I moved to Grubhub. Grubhub, the terms you care about are food delivery terms or cuisine terms that people are typing into search engines. Really, it's all about how do I create a page that says, hey, this is all you need to know about for, you know, SoMa Chinese food. Here are all the restaurants that are delivering to that neighborhood. Here's how to order. And doing that in a way where it scales for any neighborhood, any zip code, any college, any city name that someone might type into a search engine. So a lot of what we built is a backend system where I could just go in and say, Yep, we have enough restaurants in West Hollywood now. Let's turn on a page for West Hollywood. And hey, now we have enough pages, we have enough restaurants for Chinese food in West Hollywood. Let's turn on that specific page so that when we had relevant content, we could automatically show that to search engines and therefore show it to users and get new people to try us.

Got it. OK. Now, so for all these listings, I'd imagine there's competitors out there. I mean, there's like the Eat 24s of the world and all these. How do you guys differentiate your listings from the other ones? Is it user-generated content? What is it exactly?

Sure. Talking specifically related to Grubhub, it changed a little bit over time. When Grubhub started, it was actually not an online ordering platform. What the founders of Grubhub did is they actually grabbed the menus from every restaurant that delivered and then manually called up all of those restaurants to map the delivery boundaries of all of those restaurants. It was the only place on the internet where you could legitimately say, see every restaurant that delivers to you. That wasn't something that Seamless Web at the time or E24 was doing. That was the original value proposition. Then as we started converting more of those restaurants into online ordering, online ordering became an even stronger value prop. In order to differentiate there, What we really started to work on was customer service where Grubhub would take ownership of the order no matter what problem might have occurred. So if the food was late, if you didn't like the food, that sort of stuff, Grubhub would take ownership. Grubhub would make sure you had a good experience because what we realized is that people associated their experience of ordering with a particular restaurant with Grubhub. If Grubhub did everything fine but the restaurant messed up, it still was on Grubhub to fix it in the eyes of the consumer. We were the first online food delivery company to really take ownership of that customer relationship. But as you see with most startups, if you're doing something that works and you're going faster than other people, the other companies are going to respond. What we started seeing was a lot of convergence in the value props over time, where SeamlessWeb was at first focused on corporate ordering, not on the consumer. They became more consumer-focused. We were the first to be focused on customer service, and then some other services started really trying to get better at that. So it definitely becomes a race of innovation, and you can't say, well, because we're focused on this and this other company's focused on that, we're going to win, because if you start winning, then everyone else is just going to respond by adopting your tactics. So, if I could say one thing that led to us winning over the long term, it would be consistently thinking about what's the best user experience, honing in on that, and then building a scalable growth strategy around it so that someone couldn't just adopt what we're doing and usurp our lead in any one market. Got it.

Okay. Now, I mean, obviously best user experience for an online product. I mean, if you have a physical product, obviously it's the best quality, right? Okay. Now, it sounds like, I mean, for Grubhub, going back to that, I mean, you had to build up a lot of inventory around the restaurants and all that. So, can you talk to us a little bit about how you— what that process looked like? Did you have like a— did Grubhub have like a sales team or like a robocalling system where they just dialed out to everybody to kind of get this inventory?

Yeah. So, Grubhub did have a sales team. We basically had a city launch process that became very replicable, which is we would decide we'd want to launch Los Angeles and we would send a couple of people out to Los Angeles. They'd talk to restaurants for a week. They would come back with, let's say, 50 restaurants that were ready to work with us. I would then put all of those restaurants into our systems so that we could create landing pages if anyone wanted to order from those restaurants. figured out how they were structured in terms of neighborhoods and zip codes to see where we had the density to really offer a compelling product. Typically, I would tell those salespeople to be like, hey, just focus on Santa Monica and West Hollywood and we'll expand from there, or similar type of advice. Then we would go to the press in that market and say, hey, we've just launched in LA. Why don't you give your users $10 off their first order on Grubhub? and let them know about this service. We'd get certain press to bite on that, and then after 30, 60 days, we would just redirect the landing pages we would tell the press to link to, to our SEO pages so that we had the authority plus the content to rank really well if someone typed in food delivery terms related to that market. Then typically, we would just keep salespeople in the market. They would continue building relationships with restaurants, signing them up, getting them to try it. One of the things that made that fairly easy for us is that we didn't have a year-long contract. We didn't have a monthly payment model. It was 100% pay for performance, no additional fees. So literally the contract is, hey, if you get an order from us, you pay us a percentage of that order. If you don't get any orders from us, you pay us nothing. You can quit at any time. So it literally made it so easy for a restaurant to try. That was hard for them to say no, and that was a key to us being able to scale from a sales perspective. Over time, as national awareness grew of the service from a restaurant perspective, we found that we actually didn't need to send people directly to the market. We could call restaurants over the phone and they would be willing to try it. We can get them to try it just by calling them, not robocalls, actually real people, but we didn't have to necessarily fly a bunch of people. fly out to each city we were launching and station them there for months or years. We could kind of do it all from the home office.

Got it. So the way you did it, I think initially you sent people in, there's a lot of hand-to-hand combat. I mean, how long did that go on for? Because I'm just, again, people talk about scaling really quickly and I'm just trying to figure out the fastest way to do this. But it sounds like you guys just stuck with your guns. You stuck with getting people on the ground or getting people to have real conversations on the phone. Then once people started becoming more aware of it, it kind of just took off on its own?

Yeah. The whole goal of growing a marketplace city by city is building a playbook that you can replicate. If you can understand how to get the supply going, how to get the demand going in each individual market, and the way it works in one market is the same way it works in the other, you can just continue to refine it so that it's more and more efficient and less and less costly. That's all we were doing. In terms of how long we actually needed to do very manual work on the supply side, I would say that for about the first 3 years that I was there, that we were launching cities, we would send salespeople there. We would keep salespeople there permanently as employees. The other thing you need to think about here is once you've been launching cities for 3 years, The cities you're launching additionally, you're launching more of them at a time because they're smaller and because you have the process more refined. And also because they're smaller, they don't need as much hand-to-hand combat. There aren't as many restaurants to sign up, that sort of thing. So that's another element of the process. But for the first 3 years, it was basically that same sort of model. Hiring a local salesperson or flying someone in, having them meet with restaurants directly, make the sales pitch. We weren't in a rush to really change that. If it was working, we were happy to do it. I think a lot of startups are like, oh, it's not completely scalable, or this should be done entirely with technology. That's not really the Chicago school of startups. If it works, it doesn't matter how much manual effort it is, we're going to get it done.

Got it. I love it. Okay. So that's the first lesson for everybody. This is how you go from a few thousand users to millions with Grubhub. Now let's talk about Pinterest. How do you get to massive scale?

Yeah. So that was an open question as I started moving from Grubhub to Pinterest. It's like, can I grow a business that already has this many users? I don't know if the same sort of skill sets apply. But what you find is that it's a lot of the same things work getting you from Your 1,000 users to 10,000 users work getting you from 1 million users to 10 to 50 million users. It's a lot of the same tools in the toolkit. The scale of the graphs are very different, but it's a lot of the same work. We mentioned SEO. One of the things that Pinterest has is it has all these pins and it has all of these boards, which is Individual people all over the world curating what they think are the best objects on an individual topic. It's really high-quality content. Someone's taking the time to say, these things go together and they represent this specific thing that I'm passionate about. The challenge we were having is that there were images, so Google by default didn't really have any idea what the page was about. We had to work a lot with our users and with the data we're able to generate about the content to really explain This is a page about woodworking. This is a page about home decor, so that search engines really understood what's this content about. When someone searches for something, should I rank this or something else? That's been really fruitful work for us to grow. We've really went from a service built around saving to a service that now can power discovery for a large percentage of the world because we just have all the content around the internet that people think is interesting related to very specific niche topics. That was a pretty interesting transformation for the company, and it's certainly something that search engines reward us for as well. Other things that apply in a very similar way from my earlier career, things like email marketing are still very effective at bringing people back by showing them relevant content. That's something we spend a lot of effort on. Making sure that all that content is very personalized because people on Pinterest have widely diverse interests across the globe, and you know even even neighbors have very different interests. So that's something we spend a lot of time on, and we also spend a lot of time on making sure any landing page experience is as simple as possible, and is very focused on getting you to join the service or log into the service because that's when we can provide that personalized discovery experience, which we think is the really. magical experience of using Pinterest. It's not just about seeing some cool content and then moving on to the rest of the internet. It's about allowing us to really show you a personalized discovery experience and show you things you're interested in that you had no idea you would ever be interested in or would never search for.

Huh. Okay. Interesting. Let's jump a little into— dive a little into SEO. Obviously, you have With over 50 billion pins, you have at least 50 billion pages. So what are some SEO challenges you faced growing this thing?

Yeah. So I mean, you highlight one right off the bat, which is search engines are not going to crawl all 50 billion pins, right? And they're not going to crawl all 1 billion boards. So one of the things we had to do is just really figure out ourselves what are the pins and boards that are worth showing to search engines, right? How do we take a sitemap that has over 50 billion pages in it and tell Google, you know what? You really only need to look at these 500 million of them or this billion of them because the rest are very similar content or not as high-quality representations of these topics. That's something we spent a lot of effort on. I'd say the biggest thing though is at our scale is we had to build an SEO experiment framework. It's such a big service. There's so much content. In order for us to really understand, is something helping search engines understand us or hurting search engines' understanding of us, we had to do an experiment framework. The difference for an SEO experiment framework versus a typical experiment framework is that you split things by the page level instead of by the user. You know, and Optimizely works of, oh, Eric came to the service, we're gonna show him one version of the page. Casey came to the service, we're gonna show him another version of the page.

Mm-hmm.

We actually have to do that at the page type level. So in our case, we'll take 10% of the boards and we'll change something about them. We'll add more descriptive text to them. We'll change the title tag, you know, et cetera. And then we'll just see, you know, over time, do search engines provide more traffic to the enabled group or to the control group? That's had a huge impact on our understanding of how the search engines crawl us, how they think about our content, and our ability to grow our traffic from them as well.

Huh. Okay. Now, I think you're alluding to— by the way, everybody should check out this post. We'll try to link to it in the show notes. It's the A/B testing framework, the A/B SEO framework that Pinterest put together. How do you— I guess, how do you— if you had to spec this out to your engineers to build this out, I mean, at a high level, what would you say?

Yeah. It's actually not super difficult. The first step is you have to identify what are pages that you can group together that are fundamentally the same. In our case, we've talked about it. Pins are fundamentally the same type of page across the board. Boards are fundamentally the same type of page across the board. So then what you have to do is say, okay, now I need a mechanism to randomly sample a group of boards out from the rest, or randomly sample a group of pins out from the rest. And one of the things you want to be careful about is, you know, not all boards and pins are created equal, right? So you want to make sure that you're looking at topically, are they about the same thing? Because if one's seasonally you know, on trend and one is seasonally off trend, that could impact your results. And the other is just kind of general popularity, right? If one pin— if one pin is used to getting, you know, 1,000 visits a day and another pin is used to getting 1 visit a day, it's not really a similar enabled and control group. So you have to figure out the appropriate way to, you know, separate out enabled and control by the page type. Then what you want to do is be able to make a change that only You know, a search engine crawler will crawl on the enabled group and not see it on the control group. And then you have to think about, well, what, what metrics are you really trying to change here? Typically, it's you're trying to get more traffic from a search engine. So we just have, you know, a very simple graph that shows the difference in traffic between enabled and control. We run that data down to check for statistical significance. And, you know, in our case, we'll typically run the average experiment for 3 weeks or so. If we're seeing a gain in traffic, we'll ramp up to more pages and finally to 100% of those pages. And if not, you know, we'll shut that experiment down and try something else. A lot of this stuff you can do pretty manually. We did it manually at first and then started building an interface so we could, you know, keep track of this stuff a little bit easier. But it's not— it's not a supremely daunting task. Yeah. I will say that if you're only getting 100 visits from Google a day, there's no point in creating an experiment framework for SEO. You're just never going to get statistically significant data. You might as well just make a change and then see if your overall graphs go up or down. But when you're getting to the millions of visits sort of stage, it can be very helpful in seeing if you're really making the right type of progress or if you're really shooting yourself in the foot.

Well, there it is. There's the A/B testing SEO framework that everybody can literally go out there and build now. And maybe somebody can just go build a product to solve this problem. I don't know.

There's a market there.

Yep, there definitely is. And I would pay for it. Okay. So talk about some of the challenges you're facing right now of going international because obviously you have to deal with all these translations and I just would love to hear about the process.

Yeah. So international has been an interesting challenge for us. Translated into most languages over the year of 2014. So we're available in over 20 languages now. And typically when you think about going international, that's like the main step you think about is like, oh, you just need to make sure that people can actually read it. And in our case, that was in no way sufficient. The interesting thing about Pinterest is that it's all about the content that gets surfaced. The people that were using Pinterest in Japan and Germany, in a lot of these other countries before it was translated, they obviously knew how to read English and they were pinning and repinning things in English. What we found is that while we had translated the service in Japan, a Japanese user would sign up and they would just still get a ton of English content. And a very small percentage of Japanese people actually know how to read English. We really had to take a step back and reconfigure how our discovery engine works to prioritize local content and then train people that if they are pinning in their country that they shouldn't feel like they need to write it in English so that they will have an audience for that content or those sorts of things. That's been a big initiative and we've seen Pretty interesting movement in our metrics after we started localizing a lot of the feeds. And that's a process that's still ongoing for us because it's not just something you can do once and then it scales everywhere. You pretty much have to make the effort country by country. So that's been one interesting effort. Related to that, SEO is a very different challenge internationally because in the US we have all this content. It's all grouped together. Right. Outside of that, you have basically zero credit for that because none of that content is in the local language either. So really making sure that we can organize the content we have so that Google finds it versus finding all of the English content and says, oh, this is an English site. It shouldn't rank in google.fr. Those types of projects are things we really had to work on. And it's more challenging than you think at our scale. Just lots of localization work, lots of really understanding. If you have a picture of someone running, people don't run as a hobby in some cultures, right? So that just doesn't make sense. Learning all of those things. One of the things that I did that was super illuminating is I actually took a trip out to Paris and just ran through the product with a bunch of French people. You really just get to see, like, oh, they were trying to find one of their friends and the accent didn't work. Or they were searching for cake and cake means something different in French than it does in English. And you just notice that, like, wow, this is a much larger effort than we had anticipated. So we finally internalized that and started really making a lot of progress toward it. But Herculean might be too strong a word, but it's been an interesting set of work for us.

Interesting. So when you guys are doing international, just an SEO question, are you guys using like a .com and a subdirectory? Are you using a subdomain or like a different TLD?

Yeah, we are using a subdomain in this case. So it would be fr.pinterest.com. I'd say if I were doing things from scratch, I would probably use a ccTLD. So like a pinterest.fr. In our case, Pinterest was a little bit late to start going international with the service. So a lot of those domains were actually purchased already by squatters and very expensive to acquire. Having a consistent top-level domain at the country level was not an option in a lot of these countries. We have subdomains everywhere. If you're starting a startup today, as soon as you can, I would buy all of the country code versions of your brand name. It's money well spent. So I'll leave you guys with that advice. But yeah, the subdomains has been fine. It creates some individual challenges. I think the click-through rate on search engines is a little bit lower with a subdomain than it is with a country code domain. But you just have to make those trade-offs based on the environment.

Wow. OK. That sucks for you guys. Well, OK.

It's not a deal breaker in any way, but I mean, obviously you would like to have the country code if you can.

Are they trying— I mean, are those squatters trying to sell each one for like a million bucks?

I can't get into the monetary—

Let's not talk about the money. I'm just going to assume it's a million. I'm just going to throw it out there, guys. They're trying to sell it for a million bucks each. Okay. It's really interesting because I remember Chamath from Facebook said in the past when they're trying to— when they were trying to go international, some people recommended that they should just hire a bunch of smart MBAs and send them to different countries. But what they did was that they just hired, they went to those actual countries and hired people within those countries to build up their internationalization, and it worked out well for them. So just interesting anecdote.

But anyway.

You mentioned in a blog post, you talked about how to build a marketing team at a consumer technology company. Can you talk a little bit about that?

Sure. So, you know, it's one of the things that a lot of people have asked me about, and there's been a lot of confusion about, is, you know, there's some sensitivity around the term marketing in technology companies, and what is marketing, and what value does it provide. So, I was trying to provide a little bit of insight, you know, as someone that's worked in that field. Pretty much his entire career about the optimal ways to structure that sort of thing. The main thing I would provide is that there's really a couple different types of marketing. There is what I would call growth marketing, sometimes called performance marketing, where everything you do is very analytical and measurable. It includes product-focused initiatives like conversion optimization and SEO. It also includes paid marketing initiatives like paid search, Facebook ads, retargeting, that sort of stuff. That's one group of marketing that tends to be very different from the other, which is much more what I would call brand marketing. It's very focused on the message. It's focused on the positioning. It's focused on softer channels such as social media and PR, things that are less measurable, less measurable, more story-based. Those tend to be 2 very different skill sets. Ideally, you want a VP of marketing or a CMO that understands the value of both. Maybe they're stronger in one than the other, but they should represent that those are generally 2 different groups, and both are extremely valuable for a technology company to have. The earlier you are, the more you might lean on that performance or growth marketing side, and the later you are, the more you might lean on that brand marketing side. It kind of depends on what your challenges are as you grow a business, but both need to be present. They tend to be very different in their skill sets, and if you're hiring a team, you need to know that and what value each of those are going to provide.

Okay.

Do you—

I'm not sure if it was you that wrote this, but I think there's like an extensive blog post. I have to find it and put it in the show notes, but it just talks about What you just said in depth and says, you know, what skills you need for each of these positions exactly. And I think you're totally, you know, you hit the nail on the head. Okay.

So the—

going back to SEO for a little bit again, how do you sell— I remember doing this in the past. I mean, how do you sell the importance of SEO or how do you sell the importance of links to executives? Because they just don't get this stuff at the end of the day.

Ultimately, You need to prove it with data. So, I mean, this is how most technology companies are run. So, you know, marketers especially are used to a situation where marketing is well understood as a function. It is understood its value and its voice in the organization. And technology companies don't have that default respect for the discipline, so you need to prove your value.

Right.

With data. And in SEO specifically, you need to be able to find ways to do things and show the results, which is why the SEO Experiment Framework's been so helpful for us at Pinterest.

Huh.

Early on, you don't necessarily need to do an experiment to show that. You need to show, hey, I did this. A month later, I got this. Links is way more complicated to explain, and I actually have a blog post about this very topic because it is so difficult. Link building is pretty much impossible to prove the value of from an analytical perspective. So it tends to become more of a game theoretical exercise where I liken it to climate change. So if you— climate change can be happening or it cannot. If it happens, and you don't do anything about it, you're in big trouble. If it doesn't happen and you did something about it, you're fine. So what's worse? Doing something about it and finding out that it wasn't a big deal, or not doing something about it and having your business die? Typically, one is more profitable than the other. So what happens is if you decide not to link build but all of your competitors do, You have a high chance of dying, and then if you— all you need to do is link build as well to make sure you don't die. It's generally the profitable thing to do in the long run. So it's definitely a game theoretical argument. It's something that you want to make sure you have a good amount of time to go through with an executive, but widely understood to be a very important function depending on the industry, right?

Yeah.

If you don't have a lot of competitors, you can just have great content and still rank. Typically on a search engine today, there are going to be a good amount of competitors and links are an important component. We happen to be in a pretty good position at Pinterest where we get a lot of links naturally because there are, you know, it's a lot, it's user-generated content site. People link to their content. There's a lot of press about Pinterest. So we don't tend to worry too much about link building, but extremely important function for SEO in general. And if you're not going to get those links naturally, You need to find ways, not against any Google guidelines or anything, but you need to find ways to try to attract those links better if you want to rank well. It's just how it works.

Right. I think you bring up a good point. I mean, there's a spectrum, right? When you're first starting out, it's more of the manual outreach, hand-to-hand combat type of stuff. Build out the content and just let things compound. Let the links compound over time. But once you get to a certain scale, You have a lot of brand equity. The links almost just come in automatically and it's up to you to figure out what you want to do with that equity. So, you guys have obviously figured out how you guys want to pass link equity in your website and kind of maximize the value there. But I think, again, there's 2 different spectrums, right?

Absolutely.

Okay. So, we've talked a lot about SEO. You've talked about how you guys manage retention with email. I mean, what else has worked for you guys in terms of growth?

Yeah. One of the other things that's a little bit unique to Pinterest that we spend a lot of time on is working with publishers and partners. So, you know, every piece of content on Pinterest comes from somewhere else. And if publishers and partners optimize their service to drive content into Pinterest, that makes sure that, you know, people have a good experience and they're seeing the best content, but also means that people click on that content which drives traffic back to these publishers and partners. We've really worked to educate that group on how they get value from Pinterest, adding things like the Pin It button to their site. That's also a way that drives awareness and exposure for Pinterest over the long term as people see those Pinterest buttons across the internet and then they click on them and they're able to save that content and create an account. That's another way that's driven a lot of value for us from a growth perspective. That's probably the main other way. We definitely have our user base share a lot of content to Facebook, so that's a nice awareness driver and acquisition driver as well. Those would be the growth drivers that I work on. From an overall business perspective, Pinterest grew mostly via word of mouth historically. It was an invite-only service for a very long time where you had to request an invite and once you were invited, you had a certain amount of invites to bring your friends in. That was something that drove a lot of word of mouth in the early days and even today, most people come to Pinterest from one of their friends telling them about the service. But as to talking about the things that growth optimizes, it's generally search engines, Facebook, and publishers and partners are those 3 big buckets that we work on.

I love it. Okay. Let's shift the conversation to you. Before I do that, actually, one more SEO question, the last one.

You like SEO.

I do.

I do. Especially at the consumer technology level where you're at, the scale where you're at, I think it's incredible. But, you know, I mean, obviously Google's been cracking down for the past few years already. And it's made, you know, they've made SEO harder. It's just not simple to get into anymore. And, you know, I think as a result, there's just not as many people going into SEO or really learning the discipline anymore. I think, you know, you've been doing it for, you know, a decade or so. I've been doing it for a couple of years and I've managed to sneak in before Google made it tough. How do you find good SEO people?

Yeah, it's very difficult. In particular with Pinterest's case, we are operating at a scale from an SEO perspective in which the challenges that we deal with are fairly unique. There aren't a ton of services that have over 50 billion pages that are shown to Google. I've really just sought out companies that when I search I see relatively often, that I know they're operating internationally, they're operating at a very good scale, and try to build relationships with them, talk to them, see the things they're seeing, have more of that open dialogue as long as they're not competitive. That's been a value for me, and I think it's been a value to the other folks in the industry. The other, which I think is less valuable perspective to get, is to go to a bunch of the conferences on SEO. They can provide some value, but there's generally diminishing returns there because those panels and whatnot tend to talk about the same things every year.

Mm-hmm.

So I found that really seeking out people that are doing really great SEO, at other companies has been the most valuable thing to do. Even though I've worked a long time in SEO and other parts of growth, it just never hurts to get another person's perspective as to what's going on because people are encountering different problems and they might have picked up on something you didn't that's incredibly valuable to learn or will help you not waste a lot of time that you would have wasted otherwise. So that's the main way that I've done it.

Totally agree. Okay, now we can finally shift it to you. So what's one piece of advice you'd give to your 25-year-old self?

Yeah, learn to code. Yeah, more seriously, there's kind of a personal level and there's more professional level. Personal level, 25 is around that stage where Your body tends to start telling you things that you can't continue to do at the same rate. So, just really thinking about exercise, really thinking about building a healthy diet. That's the personal stuff. Professional, your career is really about the people you meet and the skills you collect. So, really focusing not as much on a title that you're shooting for or necessarily a certain role, but what are the things I'm learning today? How do I learn more than what I'm learning today? That's really the stuff that pays off down the road. So collecting as many skills as possible related to your career, just personally, that maybe aren't part of your career now, but maybe will be part of your career later. That's probably what I would tell myself. Okay.

And how do you structure your day?

I find that people are increasingly structuring my day for me. I wake up and all of a sudden I see like 5 additional meetings added to my day. No, but there's a few things. So I manage a team. So Mondays are typically— I have a bunch of one-on-ones with my team. To figure out, you know, what we're trying to get accomplished for the week, what challenges they're having, any insights I need to be thinking about. And then we also set the tone for the week with a metrics meeting where we go over the metrics for the previous week. What happened? Why did it happen? You know, what are ways that we can improve upon the good things, maybe mitigate some of the bad things that might have happened? So, you know, Monday is a lot of that setting the tone. for the week. Tuesday, we tend to do more deep dives into specific topics on the sub-teams. So growth is made up of an acquisition team, a conversion team, an engagement team, an international team, you know, and we have separate meetings for all of those things where we break down the metrics at a deeper level. We do a deep dive on one specific strategic initiative we're working on. We get up to speed on all of the individual tasks, that sort of thing. Wednesday is kind of our, you know, let's eliminate meetings and let's really hammer through large-scale things we want to do. So this will typically be where I'm doing analysis or I am writing specific docs on new projects or initiatives we want to get started, those sorts of things. And Thursday is typically a day where I'll do some outside chatting with folks like you today, get that perspective from other people, meet up with the broader product team to understand other things that are going on. So that's generally how things are structured. A lot of it is by day, and then a lot of the in-between are kind of those meetings on strategic initiatives. Making sure that we have alignment across the team and can move as fast as possible. So I'm kind of that meeting buffer, so a lot of the other members of the team can work fast and not be disrupted.

Love it. You know, it would be great if you created some type of course around how to run the growth team at a consumer technology company. I'd pay for it.

I'd pay for it too.

Okay. What's one must-read book you'd recommend to everyone?

Sure. I think I would read— if I only had one and I was working it all in something similar to the field I'm working in, it would be Thinking Fast and Slow by Daniel Kahneman. Daniel Kahneman is basically the founder of a new field called behavioral economics. He won the Nobel Prize in Economics a couple of years ago. What behavioral economics is, is a study of how people actually make decisions as opposed to traditional economics, which is about how people are supposed to make decisions. And if you're at all working on any sort of business, a big part of it is understanding how people make decisions and how they're going to react to the things that you're showing them and the things that you're telling them. And, you know, Kahneman breaks it down in a way where it's really easy to understand. It's really easy to understand That it's not just dumb people that make these decisions. The smartest people in the world make decisions like this. It's just how our brain works naturally. That would be the one book I would recommend. I definitely continue to use it even today.

I love it. The feeling I get is that you have a very close number 2. Do you want to reveal that?

A very close number 2. Okay. I really think I probably have like Give them all away.

Let's do it.

Okay. One would be Black Swan by Nassim Taleb. What Black Swan is about is really understanding that certain things are good at being measured by statistics and certain things are not. The certain things that are not good at being measured by statistics are things that can have a very Wide and most of the time unknown level of impact at the tail. At the very negative stage or the very positive stage. If you think about the difference between minimum and max height of an American male, there's a very limited scale from maybe 2 feet to 8 feet that you could be in. But wealth, You know, can go everywhere from, you know, $0 all the way to billions. There's just a much bigger scale, and honestly, this scale keeps moving further and further out. So understanding the differences in when statistics make sense to use, when it doesn't make sense to use, what happens at some of these tails— really illuminating, really helps you think about problem sets in a different way, and helps you evaluate evidence that comes in. or lack of evidence that comes into a problem in a different way. That would be a very important book for me to read. Crossing the Chasm by Geoffrey Moore.

Got it.

Yeah, seems like you're familiar with it. Really understanding how to make progress in growing a business from early adopters to the majority of a user base. And understanding where the gaps are, understanding beachhead strategies that help you do that is another great book. I feel like I'm missing a major one that would be really good to think about, but maybe I'll come back to it if I remember.

No worries. I'll come back to you because we're about to cap out over here, but I'll come back to you for that book. We'll drop it in the show notes for sure. So yeah, I mean, Casey, this has been incredible. What's the best way for people to find you online?

So I'm obviously on Pinterest at pinterest.com/onecaseman. I'm on Twitter at @onecaseman, and you can always check out my blog at caseyaccidental.com.

Awesome. Yeah, I mean, everyone, this is obviously— Pinterest is getting bigger and bigger, and if you haven't heard of it yet, you will soon. I mean, very important, I think. product that I like using as well. So Casey, thanks so much for doing this.

I remember my last book. It's The Goal by Elias Ahoo Goldratt.

Ah, the one that Jeff Bezos recommended.

And it is the best book on conversion optimization that is not about conversion optimization.

Yeah, no, I love it. I've forced people to read that book in the past. It's a good one. All right, everybody. Casey, again, thanks so much for doing this. Again, everyone check out Pinterest, and Casey, hope to have you on the show again sometime soon.

All right, thanks a lot, Eric.

Hey everyone, just a quick heads up that we're giving away an ebook called 29 Growth Hacking Quick Wins. We co-authored this book with Matan Griffel of One Month, and it'll give you a solid base on where you can create growth ideas from. So all you need to do is text QUICKTIPS to 33444. That's the word quick, Q-U-I-C-K, and tips, T-I-P-S as in sugar, to Thanks for listening to this episode of Growth Everywhere.

If you loved what you heard, be sure to head back to growtheverywhere.com for today's show notes and a ton of additional resources. But before you go, hit the subscribe button to avoid missing out on next week's value-packed interview. Enjoy the rest of your week and remember to take action, be unstoppable, and I'll And see you next week.

Take action.

continue growing.