Blog Post

One year after Facebook integration, Instagram relies on design by data

[protected-iframe id=”8160b616689cf66c42bd8724ef475e06-14960843-25766478″ info=”http://new.livestream.com/accounts/74987/events/2497095/videos/34131211/player?autoPlay=false&height=360&mute=false&width=640″ width=”640″ height=”360″ frameborder=”0″ scrolling=”no”]
Transcription details:
Date:
06-Nov-2013
Input sound file:
1006.MP3

Transcription results:
Session Name: The Little Details and the Big Data that Led to the Explosion of Instagram

Chris Albrecht
Om Malik
Kevin Systrom

Chris Albrecht 00:00
Thank you Scott. Just a quick aside before we bring out our next chat, I was reading on Skip this morning, for all you font nerds yesterday, and after watching you yesterday you are really into it. Sweden has a new font. Sweden Sans. You should go look that up, because it’s kind of neat when a country makes a new font. So we’re going to have next, we’ve got Kevin Systrom of Instagram is going to be talking with Om. Now the running joke around Om is that he likes to take pictures of the sky and if you look at his Instagram feed, sure enough this morning he uploaded yet another one living true to who he is. We appreciate that. But he’s going to be discussing the little details and the big data that led to the explosion of Instagram. Please welcome Om Malik and Kevin Systrom to the stage.

[applause]
Chris Albrecht 00:47
We got 2 seconds of an intro song.
Om Malik 00:49
Yeah I know [laughter]. And hating on my blue sky pictures, come on. Blue sky’s all about the future, and how things look beautiful in the sky and that kind of stuff.
Kevin Systrom 00:59
Om’s very sensitive, because his pictures are amazing, but we all need to go onto Om’s account right now.
Om Malik 01:06
Yes. It’s Instagram/@Om. It’s @Om also on Instagram too.
Kevin Systrom 01:12
While you’re at it you can–
Om Malik 01:14
When am I going to get on the SUL man? What would I have to do? I went from a guy who couldn’t take a straight picture to being artistic.
Kevin Systrom 01:23
So for those of you who don’t know, the SUL is short for suggested user list. And very early on at Instagram nearly two and a half years ago we added it to make sure that as people signed up, they’d actually get to see really beautiful photographers. And the unintended consequence was it became something that everyone wanted to join over time. But what’s cool is in building that list out, we’ve discovered lots of really cool community members. And built the list up over time, and Om really wants to be part of it but there’s a conflict there.
Om Malik 01:54
Some of my friends are on that list, so I’m just a little jealous of them, that’s all.
Kevin Systrom 01:58
[laughter]
Om Malik 01:59
Come on, I’m human. Sometimes. You want to tell us what’s going on at Instagram these days. You’ve been part of Facebook for a while now. So give us a quick update before we get into experience, design and all the fun stuff.
Kevin Systrom 02:13
Sure. Instagram is in a really interesting place in its life. We started with nearly 30 million users when we joined Facebook back in April, a little over a year ago. And now we’re at 150 million plus active users. So we’ve grown very, very quickly. It’s gone from a cool start up in Silicon Valley to something that affects the mainstream. And that transformation, both as a company and as an idea is something that has been both challenging for us, but also really interesting. So the addition of things like photos of you, where we let people tag their friends in photos, or the addition of video which really expands the scope of what Instagram is, and now with the addition of advertising, which is another challenge for Instagram in terms of building a business out of the asset that we’ve created. These challenges along the way are really fun things to work on as a company. And we’re now about 60 people inside of Facebook, so when we joined I think we were about 15. We’ve grown the team tremendously, although when I say we’re 65 people, people often say that’s still very, very small. But things are going well.
Om Malik 03:23
Do you get a lot of people from Facebook coming to Instagram? Or is it the other way around?
Kevin Systrom 03:29
I would say the way we’ve grown our company is a healthy mix. Part of our management team, Emily White, who’s in charge of business operations, she was working on business development on Facebook before she joined the Instagram team at Facebook. Peter Dang who is in charge of product management was also working on messenger at Facebook. So one of the awesome things about Instagram joining Facebook actually is being able to look around for talent that has applicable skills for what we’re trying to tackle. Because you forget that – I mean you don’t forget, it’s pretty obvious, but Facebook had to grow very, very quickly, and faced a lot of the similar challenges that Instagram’s going to face going forward, and I think that what’s awesome is that we can actually leverage the knowledge that has been built up over the years in the form of these people going forward.
Om Malik 04:16
Just don’t hire all these designers and put them in a room to do nothing. Just like Facebook does.
Kevin Systrom 04:23
I think that’s harsh [chuckle]. I’m not just saying that because I have a boss now.
Om Malik 04:30
Who?
Kevin Systrom 04:31
Mark.
Om Malik 04:32
He’s your boss?
Kevin Systrom 04:32
Yeah.
Om Malik 04:33
Since when?
Kevin Systrom 04:34
Since always. But I have to say I actually think that some of the projects – many of the projects Facebook is working on, are not only world changing, but also touch so many people in the world that they require lots of designers, they require lots of product managers. And that’s one of the challenges of growing your company. And we’re looking at this as Facebook as well. As we grow the company in terms of number of people working there, but also the scope of problems we’re trying to tackle, like internet.org, I think that it’s challenging to go from being a small company to a large company. And that transition is something that any world changing company has to go through and face.
Om Malik 05:17
So what about your ads; you launched them a couple of weeks back, and a lot of negative feedback but I probably was the only one that thought it was the right way to do it. How are they going? Are the ads working out for your guys?
Kevin Systrom 05:31
Yeah. I think there are a lot of different ways of defining working out. Are they making us hundreds of millions of dollars per day? No, but that was the goal. We announced to the community we were going to take it slow, do it the right way, figure out who the right advertisers are. So we measure, is it going well, by how is the roll out going? And how can we make this transition? Again, I’ve used the word transition a lot in the last five minutes, but it really is a transition from being a startup that is trying to become a business, to actually becoming a business. And I think it’s going really well. In fact, over 5% of the impressions led to likes on some of these ads that we’ve run. And that’s pretty tremendous considering that most of the ads we see on the internet we basically ignore. And then in terms of negative feedback, I think that’s to be expected. But one of the philosophies I had very early on when we were building this product was that ads on Instagram should feel like they should be on Instagram. And by the way the community should have a voice. So I think it’s absolutely fine if users want to be able to communicate with us. It’s one of our themes at the company by the way – is community first. And we always want to make sure that the community has a voice and tells us what they love and what they don’t love and what they want us to have changed, and what they want more of. So if you look at one of the recent ads that Michael Kors ran actually, many of the comments are actually very positive. Asking where they can buy the watch, asking what watch it is, friends ccing friends about the watch. That’s what I want comments to become on Instagram with ads, a way of communicating not only with the brand, but also sharing with your friends.
Om Malik 07:08
So, in terms of ads, how are you going to scale this? It works in a controlled environment, so how are you guys going to make it the Facebook style – mega billion dollar business?
Kevin Systrom 07:23
I think one of the questions I get asked often is we tend to harp on the fact that Instagram ads are going to be beautiful. And everyone says, yeah right that works great for three ads. But when you get up to thousands of ads that are running concurrently with all different targeting and different people, where does that go? Does it become less beautiful over time? Where are the controls in place? And honestly, we just need to face that challenge as we grow. But one of the commitments we’ve made to the community is we want to keep the standard really, really high. And the reason why we’ve picked the specific advertisers we decided to start out with is because; a – they were great on Instagram to begin with, but b – they were committed to producing great content that would feel at home on Instagram. So that’s been inspiring to see that come true. So General Electric, Michael Kors, Ben and Jerry’s, PayPal, you name it, any of these brands are doing wonderful things on Instagram already. And I think we’re just going to continue that. And that’s one of the challenges of building this business. But I also think it’s the opportunity in front of us.
Om Malik 08:22
I hope you guys make it work. Let’s talk about this stuff here – I’m very interested in knowing more about the Instagram experience. The design, and the data driven environment you guys have created. So let’s talk a little bit about that. What was the experience philosophy behind Instagram when you guys started it? What were you trying to create on day one?
Kevin Systrom 08:49
On day one we were actually working on a different service called Burbn at the time. And it wasn’t spelled Bourbon, it was Burbn, and it was a check-in app, that also let you post photos. What we realized very quickly was that people loved posting photos, to be able to communicate very easily. So I talk about Instagram not as a photography company, but instead as a communications company. I believe every photo that gets posted to Instagram is not necessarily a photograph, or a video is not necessarily a video, it’s a message. You’re trying to send a message to the world. You’re not posting into the ether where it sits on some drive somewhere to be discovered ten years from now. You’re posting it because people are going to see it and react to it. So early on we decided that we’re going to look at photos as messages. Videos as messages. Media as messages. And solve people’s problems. Allow it to be shared beautifully. Allow it to be shared quickly. And allow it to be shared broadly. And those three problems were the basis of our philosophy in designing Instagram very early on, and honestly that hasn’t really changed. We’ve still focused on making photos beautiful, making it easy to share, making it easy and quick to share as well. So those are the philosophies we focused on very early. But the fascinating thing that we’ve been talking about is the balance between hunch-driven design and data-driven design. Early on we were very hunch-driven. As we’ve scaled, we’ve started to make data-driven decisions. Why? Because it’s really easy when you have a hundred users to go talk to those hundred users, and ask them what they want. It’s really difficult when you have 150 million people in many different languages, half of whom are outside of the United States, to understand what people want and how they’re using the product. So that transition from being a hunch-driven design to data-driven design, is one we’re making slowly, but making sure to retain some of those hunches that built the products that we built very early on. I like to say that you can’t AB test your way to success. You can take a great idea and AB test it and optimize it, but you can’t design the best product by starting with an AB test.
Om Malik 11:01
Right. Do you often think back to the last two and a half, or two plus years, and how fast Instagram has grown? Does it surprise you? Do you wake up, come on, couldn’t be happening.
Kevin Systrom 11:15
It does surprise me. But it surprises me because no one expected that mobile apps could grow this quickly. Look around at the apps that are the biggest today. The Snapchats, the Instagrams, the whatever mobile app ends up being big next month. These things weren’t around a couple of years ago. And for the first time in the history of the internet, do we have growth at such a clip, that it basically means next month you may have another really large service, either to deal with or work on, and that’s unprecedented in the history of the internet. And I think as consumers we have to decide how are we going to spend our time.
Om Malik 11:57
Right. Where does the attention get siphoned off for the growth of Instagram and Snapchat, and whatever comes next? It seems like somebody is going to lose. There is only 24 hours in a day people can spend on these things.
Kevin Systrom 12:15
I think what we’ve seen as a transition to mobile apps, being much more single feature focused, so you don’t end up with the broad operating system like sites, where it has a ton of features, and it’s got messaging, but it’s also got uploads, but it’s also got a community message board. Mobile apps don’t really have the space to do that. And because of that we’ve switched to a world where – basically each of these apps are utilities in their own way, and they all own their own little space of behavior and interaction. And we typically share graphs in between the apps. And I think one of the things Facebook did very early on that’s very smart is realize that the one thing that every app would need is this social graph. And that’s why I think Facebook has become so important in the mobile world. Especially for mobile developers like ourselves. But going forward, it’s not about siphoning off attention, because I don’t think there’s going to be an Instagram replacement. I don’t think there’s going to be a Twitter replacement, or a Facebook replacement. I think that people will decide how they spend their time between these different icons on their own screens.
Om Malik 13:26
So what’s the growth challenge? Your company’s gone from zero to almost 200 million people in just over two years. You’ve seen Snapchat already over 100 million people. At least that’s what I’ve read. And you start to see that, and you see a company like Twitter, Face, grow challenges. Facebook is at almost a billion people now. So where do you find growth for all of these communities? Is there a glass ceiling for social platforms in that sense?
Kevin Systrom 14:00
I don’t think that there is a glass ceiling for social platforms. I think that our sustained growth shows that. All the different services you’re talking about are growing quickly, but the biggest challenge I actually think is, when you were based on the web and grew very large on the web, making the transition to mobile. And I think it’s something that Facebook has done very, very well. That being said, going forward I think it’s a challenge for many of the services that started out as web browser companies, now moving onto an app. And I think that starting in an app– think about Instagram, I actually don’t think it would have been successful had we started in the browser. Who wants to sit around uploading single images one at a time that they took ten hours ago, when the whole point of Instagram is to tell people what you’re doing now. The whole space we created wasn’t photo sharing. It was in the now photo sharing. It was in the now communication, through these photos. So there is a challenge in terms of growth eventually, because I think you start getting into, well there’s a certain number of iPhone in the world, although it’s growing very, very quickly, you go to android and you realize it’s very fragmented and you have to support many different systems. So that tends to be our biggest challenge. Is making sure our app works consistently across all of these platforms. Including now Windows Phone, and making sure that we can tap the larger markets. In many ways, all of the services are riding the wave of smartphone adoption, and I don’t think that’s slowing down any time soon.
Om Malik 15:33
Let’s talk about your future. What does the future hold for Instagram? In the last two years, what you have learned and how did it shape the future?
Kevin Systrom 15:43
The hardest thing for us going forward is figuring out how to take the fact that we get over 55 million photos per day, that come in from all people all over the world, describing what’s going on in the world right now in a public way. Figuring out how do we surface live events overseas? How do we tell you that there’s a riot happening in London? How do we let you tune in to the End of the World series? If Instagram is about communicating what’s happening in the world right now visually, how do we let you tune into that? How do we let you discover what interests you? If you’re into fly fishing like my friend is, he actually follows all these fly fishers on Instagram – and by the way video has transformed that entire world on Instagram – but how do you tune in and find those people with your niche interest? How do you discover the things on Instagram that are going to keep you coming back and stay interested? I think that’s a challenge because today we have basically hash tags, and hash tags search, and users and users search, but it’s very difficult to find the accounts you want to follow, the content you want to discover without being really good at searching through hash tags. And I think that’s a really tough place to be as a service that’s becoming more and more mainstream. Where hash tags aren’t exactly the way my mom or many people’s parents want to search the Instagram ecosystem.
Om Malik 17:10
There’s a lot of ambient data on the phone these days, there’s location data, pictures have a lot more metadata embedded into it, how do you plan to use all that and create the next Instagram experience? And the reason I ask you that, is that there’s a lot of times I try to figure out where a picture is taken and never get the context, and I have to click, go to a map, to see, oh this is in Copenhagen. Like MGs in Copenhagen right now. So I didn’t know what that place meant. So how do you create that visual experience, which again fits into the philosophy of having the Instagram narrative? You have the communication, you have the photographs, the beauty, all of those things but still controlled by data and ambient information like location.
Kevin Systrom 17:58
I think location is one of the key parts of a mocking and new Instagram experience. Although in the world that we live in, I think people are pretty reluctant to give location information. And I think rightfully so. There’s been a lot of headlines in the last year about where is our data going, who’s looking at it and where is it going to live? What we need to do is figure out how we balance the idea of privacy around location, but also the benefit of providing location data. Because I’m sure all of us would love to see what’s happening half way across the world as it’s happening. But at the same time that requires users of Instagram and other services to provide location data. So I think there’s a balance, and there’s probably a solution there. The other part is proximity. I actually think knowing that, if we’re out for a coffee and you take a photo of that coffee, that I was with you, and that becomes a story, so I can go back into my history and see all the photos that I was near when it was taken. Because you don’t have to be the first person shooter, as it were, you need to be near the other photos, or near the person taking the photo, because when that happens you start to build this collection throughout your life of moments you were near. And I think proximity and location are both similar opportunities and I think that we’re just not at a point yet with proximity. Data that we’re able to do as well as we want to do. But I think that that’s going to be unlocked easily in the next few years.
Om Malik 19:32
Are you guys going to add some more tags, apart from people and…?
Kevin Systrom 19:37
One of the ways we discussed tags very early on, especially with people tags, is that we didn’t call it people tags explicitly, because we wanted to leave the opportunity open in the future to be able to tag other things. It might be really interesting for a business to be able to tag a product so that – like I discussed in the Michael Kors example – imagine you see a watch and you say well what is that watch?, you tap on it and you actually get to see what the watch is. That’s very future thinking. I don’t think that’s a reality any time soon, because I don’t think consumers necessarily want to sit around tagging their photos with products. But I do think that it’s interesting to think about a metadata layer on top of photos, that lets you create what’s inside that photo. Whether it’s a product, whether it’s the place you’re at. Whether it’s the person you’re near or you’re with. We need to figure out how those things evolve and how they end up becoming part of the Instagram experience, while keeping it simple. Because everyone loves how simple Instagram is.
Om Malik 20:43
So how are you going to do this whole new view into the world, the real time view into the world, give people all this data, how are you guys planning to do it? What is the thinking? How are you thinking about simplifying the data?
Kevin Systrom 21:01
I think that the best case scenario is where – like you said, you have this ambient data, and you are aware of what it’s collecting, and you’re aware of how it’s getting used – I think that’s the biggest challenge. And I think what we need to do is figure out how to present that in a way that makes it really clear, and by the way provides user benefit. I don’t think just telling everyone, hey let’s all of a sudden GO tag all of our photos, makes a lot of sense. Nor does it provide an immediate benefit to me as a user. So the challenge is in displaying what the benefit is for you as a user, and also discovering how many people share, how do we understand how to display that information to you, and how do we detect things. If a bunch of photos are getting taken overseas in one place, how do we detect that? I think that’s a science we haven’t quite figured out yet, but it seems very, very solvable.
Om Malik 21:57
Yeah that’s an interesting point, how do you balance data and the human aspect of technology? Where we end up in most applications is that they’re very data-centric, and there isn’t enough emotion. How do you guys do it right now, and how do you plan to do it, especially with so much more data coming into the system now?
Kevin Systrom 22:22
Well Instagram is an emotional experience. Seeing videos of a surfer from GoPro or seeing a photo from the top of the Eiffel tower, these are emotional experiences that we have. Or whether it’s a friend’s dog, or a baby’s first steps, these are all emotional experiences, and the great thing about visual data, is that it is emotional. And this other data, the metadata on top, we have to figure out how that makes these experiences even more emotional. How would you have seen something that you wouldn’t have seen otherwise? I think that’s really the answer.
Om Malik 22:55
Right. Somebody sitting in the audience right now, how should they think about it? What advice would you give to them when they’re thinking about data?
Kevin Systrom 23:08
The advice I would give when it comes to data, is not to think about data that just benefits the company, think about how it benefits the consumer. Think about when a consumer is sitting there deciding whether or not to give you data, whether it’s a photo, a video, a GO tag, a caption, any of this that we call data just because it’s stored on a disc, think about what the benefit it to the consumer, and think about how the consumer thinks about it. Not about how the company thinks about it, not what the benefit is that you’re going to describe in your management meeting is, but rather when the consumer is sitting there deciding to either participate or not, what is the cost benefit analysis? What am I worried about when I’m providing this data? What am I thinking about? Put yourself in the consumer’s shoes, rather than in the manager’s shoes, that’s where you end up building the right experience when it comes to data collection, and also data participation.
Om Malik 24:01
Right. It seems like you guys started out with the narrative of Instagram, saying this is what we are, this is our story and we’re going to stick to it until the very end. You made scenario changes here and there, is that the right way to think about creating a wonderfully designed experience?
Kevin Systrom 24:21
The best products in the world have a point of view. The worst products in the world have no point of view. They’re scattered. They don’t really have any central thesis. With Instagram I tell my team often, when we have designer reviews, and we talk about the products we’re working on right now, we should have a very strict and very serious point of view. Not in the sense that it’s too serious. But that it’s scoped. That we have conviction for what we’re building. Because honestly, like a lot of the decisions you make with a product, one way or the other, it’s not that one completely works and one completely doesn’t. Yes that’s true if you make terrible decisions. But I’m assuming that people who work on products in Silicon Valley typically make decisions within a domain – these are decent decisions. And what I tell my team is, within that set of decisions you’re making, or the solutions you’re deciding on, you’re not going to design a bad product or a good product, you’ll design a different product. And you just have to decide what product you want to build. Is it more open? Does it feel more intimate? Is it real time, does it feel slower? And is that a good thing? Because you can delay decisions? All of these different variables just design different products. It’s not a better product or a worse product, you just have to have a point of view, and then run with it.
Om Malik 25:50
It seems like you’re taking a more adjective-driven approach to design and creation is the way of thinking at Instagram it seems.
Kevin Systrom 25:58
The thing I was taught in high school and the reason why I come back to this is because I often didn’t do it. You start your essay with a topic sentence – what’s your thesis? And when you think about product design often you need to start with a product thesis. Instead of just rambling on for paragraphs, and I realize I’m drawing a tenuous connection here, but when you’re building products you can’t just ramble on, you need to have a central thesis for what this thing’s for, how it benefits the consumer, how the consumer should think about it, and what you’re trying to prove to the consumer the benefit is. And if you start with that thesis and then support it, in the following paragraphs, then you’ve been able to write a persuasive essay. It’s the same with design – you need to persuade people to use it, you need to persuade people that it should exist, and that it’s going to benefit them. So everything should start with that central thesis, whether it’s deciding what the main problems you’re going to solve for the user are, or the point of view. You just need to understand that bland products will lead to bland outcomes. And I think that sometimes the more divisive products, things like the app that a lot of people talk about Frontback – why would you want to take a front picture and a back picture at the same time? Well it turns out that actually that’s really interesting, and actually people love doing that because it effectively allows you to take a selfie without just taking a selfie. And people love that. And there’s a whole market for it. So having a point of view and being a little out there, often leads to better products.
Om Malik 27:31
Are you going to introduce a Frontback-like feature in Instagram?
Kevin Systrom 27:35
It’s not on the road map any time soon.
Om Malik 27:36
Alright, just making sure.
Kevin Systrom 27:38
He’s waiting for the headline.
Om Malik 27:40
No, I think the headline that I’m trying to get to is that your first comment about Instagram being about communication, and my opening remarks this morning, I think we’ve entered into the age of visual communication in a very big way. Whether it is the success of you guys or Snapchat, even Twitter. What do you think happens to guys like Twitter?
Kevin Systrom 28:09
In terms of visual communication?
Om Malik 28:10
No, from their core as essentially sharing short messages, 140 characters.
Kevin Systrom 28:17
I think they’ve shown tremendous growth and tremendous value for the world in terms of distributing information really quickly. Whether it’s news or links. I think there’s a market for that. I think there will continue to be.
Om Malik 28:29
Can we do something about Instagram on Twitter – do you want me to broker something?
Kevin Systrom 28:35
It’s a much more complicated conversation [laughter].
Om Malik 28:37
I know it is, but some of us would like more people to see our pictures, and Twitter– Thank you Kevin for making time, and more good luck at Facebook and at Instagram, and keep creating those magical moments.
Kevin Systrom 28:51
I really appreciate it. Thank you.
Om Malik 28:52
Take care.
Kevin Systrom 28:52
Cheers.

[applause]