Pinterest’s Third Way


Ben Silbermann and company have built tech’s most misunderstood platform. 2018 could be the year it breaks out with a model markedly distinct from its Valley cousins.

Pinterest CEO Ben Silbermann

Like Facebook, Twitter, YouTube, Instagram, and Snapchat, Pinterest is a digital platform with hundreds of millions of users and an advertising-based business model. It’s also a “decacorn” — one of very few Valley companies with a massive ($12 billion+) valuation and an enviable list of highly regarded investors.

But once you dig into the company, these similarities quickly end.

I’ve recently spent a fair bit of time with several Pinterest execs, including President Tim Kendall, who is leaving the platform after five years, and Pinterest CEO and co-founder Ben Silbermann (my interview with Silbermann is below). To grow into its daunting valuation, the company must build a billion-plus dollar top line — a feat rarely achieved these days, given Google and Facebook’s stranglehold on digital advertising. But the company is half way there: While it doesn’t publicly comment on its numbers, sources tell me Pinterest hit its goal this year of nearly $500 million, and is on track to grow that to $700 million or more in 2018.

All this raises an obvious question: Is Pinterest chasing the same dopamine-driven ad model as its Valley kin — a model recently called into question by former tech executives, not to mention consternated Congresspeople?

The short answer: No.

The more time I spend considering Pinterest’s core DNA, its founders, its platform mechanics, and its advertising model, the more I sense a new approach taking hold. As Silbermann will tell you — over and over — Pinterest is a mission-driven company. And that mission — to help you discover and do what you love — is all about you. Not “us,” not “we,” not “them.” But you. What you want to do. What inspires you. What you can do with your time to make your life a little (or a lot) better.

Pinterest is absolutely not a social media platform. It’s far more like search in the early days — but with all the trimmings and chrome of a modern digital platform. It’s not about sharing, validation, likes or Instabragging. It’s about getting shit done. For yourself. As Silbermann told me in the interview below, it’s not about being entertained (Facebook), or getting an answer to a specific question (Google). People come to Pinterest with an itch to scratch, a task to be done. They want to figure out what to make for dinner, or how to redo a bathroom, or plan a wedding. And they build intricate maps of those intentions through their Pinterest boards. In the past few years, Pinterest has gotten quite good at helping their users discover new ideas to add to those boards.

And of course, if you’re an advertiser looking to offer a related product or service, well, somewhere in all that ideation is one hell of a selling opportunity.

Pinterest is maddeningly hard to get your head around if you’ve spent as much time as I have thinking about search and social. But the light went off when I started to think about its scale (200 million+ users) and the aggregate data those users create every day. Pinterest calls part of this data the “taste graph,” a concept that perhaps failed to capture the world’s imagination as yet, given all the graphs that have preceded it (Twitter’s interest graph, Facebook’s social graph, etc). But the taste graph is a powerful ally if your core business is helping people discover actionable ideas — ideas they can go do. Pinterest drives its business by that one metric — does the service help people get things done? Last year it implemented a “Tried It” button that has become a much loved feature of the site. And unlike most “social” sites (YouTube, Insta, Twitter), trolls don’t thrive on Pinterest. Insulting comments about a Pinterest project tend to get shouted down by community members who ask “Well, did you try to do it?!”

It’s a refreshing sensibility in a digital world overwhelmed with toxic externalities. While they can certainly still develop, it seems that present day Pinterest has no negative externalities. People actually feel better after a session on Pinterest, company executives have told me. On those other platforms? Not so much.

2018 will be the year that Pinterest will prove if it’s worth that $12 billion valuation. It’s worth noting that the fellow now in charge of its ad products is Jon Alferness, a longtime Google Adwords exec who ran both Travel and Shopping for the search giant. As Silbermann told me toward the end of our interview, Pinterest’s model reminds him of early Adwords, back when all parties to the transaction — the user, the platform, and the advertiser — felt good about the value exchanged (Silbermann worked on Adsense for two years earlier in his career). Can Pinterest make that happen at scale in a world dominated by the Dread Duopoly of Google and Facebook, not to mention Amazon, which is looking to undermine them both?

I think the company has the best shot of any at doing just that. Silbermann will join me on stage at the Shift Forum for an interview in late February. I hope to see you there!

John Battelle: The narrative around technology companies over the past year or so has been: “Well, it’s over. Facebook and Google won, and there’s no interesting up‑and‑coming, fast growth, good for the world services left. If there are, they’re all going to get bought by one of those guys.” I hope you don’t get bought by one of those guys in the next two months.

Ben Silbermann: I don’t think so.

Well if you do, I guess we can hear you explain that at the Forum! Given Pinterest is not always well understood, what do you hear about the service that really irks you, perhaps because it reflects a misunderstanding of the service at some fundamental level?

One thing we hear a lot from people who have never used Pinterest before is “I don’t need another service to keep in touch with my friends.”

I think that’s a really common misconception if you haven’t used the service. Pinterest is very much like a personal utility. People use it as a personal tool to get ideas to do things in their life. While there are people that use the service (for other things), the goal is not to broadcast to them, or to get them to like your stuff.

That one bothers me, mainly because it’s a barrier to people trying it out.

They think it’s another social media site, and they don’t need another social media site?

Yeah, they don’t feel the need to have another way to communicate with their friends or their family. They feel taken care of there.

Is a corollary of that is like, “I don’t need another place to take, or save, or share photos or images, because I’ve got Instagram”?

For us the first one is the one that we spend a lot of time clearing up. It gets back to how people use the product. They use it literally to make plans, and to think about their future in all these very personal ways. There’s really nothing about it where the goal is to broadcast to other people… They use it to save ideas for themselves, but they don’t really increase their usage in order to build an audience or anything. It’s just a personal tool for themselves.

That said, you do have the equivalent of influencers, right?

We do, and they’re businesses, they’re professionals that have their content really well represented. For the average person, what we always think about is if you put up a Facebook or an Instagram picture and nobody likes it, it’s kind of weird, like you didn’t do it right. Whereas on Pinterest, if you put up something you want, or an outfit you want to try out, or a place you want to visit, you’re not waiting for people to respond to it to know whether or not you should have saved it to Pinterest. You’re saving it there for yourself.

A hotly discussed topic in the Valley and beyond is the system of feedback loops, engagement metrics, and unconscious behaviors that many claim negatively drive advertising-based businesses like YouTube, Facebook, Instagram and Snapchat. But Pinterest is not driven by “time on site,” is it? What would you say the metrics are that do drive Pinterest?

The mission of the company is to help you discover and do what you love. It’s been very consistent for the last six, seven years. We try to figure out how many people are finding things, discovering things, and then trying to go out and do them in real life. If that’s a recipe, it would mean you go and cook it. If it’s a product, it might mean you go buy it. If it’s a home project, you might finally get it done.

That’s the North Star of where we want to get people. Our theory is that the more people who actually do things, those are the people that become the evangelists for the service. That’s born out in the research that we do, in the data. We’re always trying to focus on that, rather than just focusing on aggregate time spent.

But like Facebook and other social sites, Pinterest has advertising as its core business model. So if it’s not time on site — which is Facebook’s driver, then what’s on offer for advertisers? What’s the unique differentiation of the advertising platform compared to other services?

I think there are two things that are very attractive to advertisers that are more unique. There’s one that’s less unique, but important. The one that’s less unique that’s important is just there’s an enormous audience of millennials in a very attractive demographic in the United States, and globally.

I’d say of the two things that are unique, one is what people are doing there. They’re looking for something they want to do, so there’s intent, but they haven’t made up their mind exactly what they’re going to do. They’re not just there being entertained (ed — as with Facebook), nor do they have that long tail search query that they’ve put in (ed — as with Google).

I think for companies who are looking to find new customers or introduce existing customers to something that those people didn’t know was out there, that’s a pretty ideal environment to reach people in. You can really get people who are in consideration mode, but haven’t fully made up their mind.

The other thing that’s unique is that the content that these businesses provide, it’s put into an environment that’s really native for that. People are there. A lot of the content is already visually beautiful. It’s meant to be inspiring.

That’s become really important to a lot of these advertisers, especially in the last year, where they want to be in an environment where they feel native, they’re not intruding, and it’s a safe place for them to advertise.

Years ago we talked about wanting to have a service that could counterbalance what was happening in social, which there’s been a lot of value, but it’s really focused on other people.

That safety thing is, obviously, on a lot of minds. With all the fake pages, the fake news, the YouTube blow ups. I know all platforms have spam, but Pinterest doesn’t have the same economic incentives to create that kind of spam, does it?

I’m trying to be the most intellectually honest as I can. I think every platform has spam. We’ve tried to build systems that make sure the environment’s good for people to do what the mission of the company is. It’s been important from pretty early days.

Let’s talk about the data that your platform creates. Can you lay out the kind of information that Pinterest has and the scale of it?

One thing that’s unique is that Pinterest has very, very granular data about what people aspire to do in the future. That’s important. That’s in a lot of ways the most important thing for businesses, because they’re selling their goods and services in the future, not in the past. It’s highly structured. I think the second thing is there’s a lot of data on aesthetic and subjective tastes and preferences.

We talked about it earlier this year — this idea of a taste graph, given that you like this style or this category, what are the other things you’re likely to enjoy? Again, I think that makes a lot of sense for people looking to sell things. Those two things are pretty unique to Pinterest.

Computer vision is starting to take off, and you also have a unique platform in that space. How is Lens going and what have you learned from rolling that out?

Lens is a project that represents kind of a broader investment in computer vision. We think computer vision’s incredibly important. We’ve talked about it at length. I think that vision is where speech was a few years ago, where it’s just improving substantially.

We do three things with computer vision. We try to understand what are the aesthetic qualities of a product or a service, so we can do better recommendations, in general. We want to be able to look inside an image that has multiple items, zoom in on part of it and computationally say, “Hey, we think it’s this type of object. Here’s where you can find something similar.”

Then, eventually, we want to make the camera tool that you can query the world around you, and computer vision sort of powers all three of those things. The third one, which is like what we call Lens is definitely the most experimental. It’s like the most sci‑fi, but we wanted to put something out there early, so we started learning about those experiences. … I just think computer vision is going to be one of these really fundamental technologies.

But Pinterest is an upstart compared to Google, Microsoft, or any large tech company that has 10 to 100 times the developers and engineers working on computer vision. How do you the approach it differently?

Pinterest has two principles that guide it. One is that computer vision should be used towards the service of helping people discover and do things that they love. We try to keep our eye on what are the use cases that are going to be the most useful for people. Then I think the second thing is Pinterest does have a very rich data set, that taste graph, where people are grouping things together based upon their tastes and preferences. That means when we train these computer vision models, we have a lot of great raw data to work with.

The three things you would need for any computer vision problem is of course you need great engineers, but you need really rich structured data and you need strong computational power. We can buy computational power in the cloud. We have a team, a good team. It’s a smaller team, of course, than Google. And we have good structured data. We have a really good place to play.

Are you struggling to find and hire the people you wish you could? It’s a very competitive market for this kind of talent.

I think every tech company is working hard to bring in more and more great talent. The benefit and the reason we’re able to attract good people is this idea of having out‑sized impact, given the stage of the company. You’re always trying to differentiate on what’s special.

Let’s talk about why a senior computer vision engineer might choose to work with you. Clearly one reason is the stage of the company, which is exciting for anyone. It’s pre‑IPO. It’s not a big already public company. But the other thing is the way the company is run, the culture of the company, the sensibility of the company. Can you describe that difference?

You already mentioned some of what we lead with, which is the mission of the company. Even before I get to the culture, I really think the mission of the company in helping people discover something for themselves, execute on that, feel creative and try something new, I do think that’s a rare space that we’re going after.

While there are companies that are kind of adjacently trying to get that, I think Pinterest is the only company that’s 100-percent devoted itself to you, your personal aspirations. So that’s a pretty big deal.

Then there’s the scale we’re at. There’s well over 200 million users. There are just a little bit over 1,000 employees. We’re working on a really big scale on really challenging technology problems.

Culturally, one of the things that’s special is there’s a real obsession over using whatever tools we have at our disposal to solve real customer problems. Of course, there’s big technology tools, but we ask people to work really tightly with very good designers, very good writers, with very good researchers. I think that multi‑disciplinary approach is different than what I’ve seen in other technology companies.

Would it be fair to say that Pinterest is actually something of an antidote to what ails a lot of tech’s engagement with culture today? This past year, a lot of people said “Oh my God. I didn’t realize that being on my phone all day is bad for the world.” Is there’s a feeling at Pinterest that you’re doing something that’s more, I don’t know, healthy?

Years ago we talked about wanting to have a service that could counterbalance what was happening in social, which there’s been a lot of value, but it’s really focused on other people. A lot of our users talk about this phenomena of “me time” — I want to think about myself, my aspirations, what I care about, and Pinterest being one of the very few tools to do that.

From a brand perspective, if you look at where our early users came from, they weren’t from the stereotypical places where tech is adopted early. I think from a mission standpoint, there’s a thought of Pinterest being a counterbalance for people in their lives.

Does this create a sense of pride within your employees? You try to get people off the site and doing things. That’s a goal of your company. You’ve got this relatively new “Tried It” feature…

Tried it. Yeah.

Is that an important metric?

Yeah, all I talk about with the company every week is whether we’re getting closer or farther away from our mission as a company. That is taken really seriously. We’ve talked for years about how using attention as a singular output metric is not a good way to run a business. It’s just a bad way to run a business.

We’ve talked for years about how using attention as a singular output metric is not a good way to run a business. It’s just a bad way to run a business.

I’m hesitant to make a broad statement about companies that are as large and diversified as Google and Facebook. But I’m glad that the conversation about what does the quality of that time mean, is being had. It’s a credit to those companies that are actually talking about it now after the full, what, year and a half of like hard denial. It’s great that they’re talking now, because if you talk about stuff, then you can solve it, right? [laughs] I think that’s good.

Although I think they might be in a bit of their own trap. When I think about the growth they have to maintain in order to justify their valuations, it’s really hard to walk back the attention-based advertising model they’re in, right?


What are things that you wished advertisers understood about the service, even the ones that buy from you? Are there things that you wished they understood about how to succeed on Pinterest?

The advertisers that have seen the most success have measured it based upon whether Pinterest helps them drive new sales. They’ve had to invest to measure Pinterest slightly differently than they measure Google, which is very last click.

The ones that have jumped on that early, and said, “Hey, we think that we need other channels to get new sales,” versus driving same number of sales on a shorter time window, have seen really, really good success. It’s really that simple. It’s like if there are like tens of millions of your customers in a place, and they’re planning their life activities that would involve your product, it’s like do you want to be present in that process or not? The ones that have started have had really good success so far.

We have a lot of case studies that are coming out about that. I’ve been really excited because they’re spending more quarter‑on‑quarter, year‑on‑year, and they’re doing that because they’re getting really good measurable results.

You do have to train them to understand a different value process, right?

Yeah, the default is that most people are spending all of their money on Google and Facebook, so to introduce a third option requires a little bit of work, and our job is we’re working hard to make it as little incremental work for them to see the benefits as possible.

Do you see a clear path to a really big business? Do you see any other revenue sources besides advertising?

I think there are lots of possibilities, but the reason we’re focused on advertising is because if it works, it should make the service better on both sides.

For people out there to plan their future out, part of that is finding things they can buy and do. Businesses are there to have their stuff discovered. If it works, then the flywheel turns both sides of it…

Sort of like early AdWords?


I see a pretty strong comparison to early AdWords. The ads drove value to all parties, not just the platform.

I worked in Adsense. I think we have an alignment in …what people are there to do, and the objectives of the advertiser. I think a lot of the best advertising products are built in that model.

I, for one, hope you succeed.

Count me in. I’m a plus one on that.

Leave a Reply