Power of Tech Platforms

Being a platform on which other businesses can thrive has long been a holy grail for Net companies. Companies that want to loom large in the future almost have to control some sort of platform, ideally the center of an entire ecosystem devoted to the company’s survival. How should platform thinking inform the strategy of every business? How does it apply to industries not traditionally considered “tech”? How many platforms can the world stand?
Read the full transcript below. (Transcript by Realtime Transcriptions.)
Kirkpatrick:  The next session is going to talk about platforms, because it's actually been mentioned throughout our discussions on numerous occasions how important being a platform is becoming in business. So we thought we would examine that directly. That's what this next session talks about and how it might apply to as many different industries and sectors as possible. But certainly we know it applies deeply in tech, and we have some really smart tech people here to talk about it.
And moderating is Katie Benner, a long-time colleague of mine when I was at Fortune who is now at a still unnamed company where she's blogging under the Jessicalesson.com URL, because Jessica, who was one of the top writers at the Wall Street Journal, recently left to start a new media company, hired Katy as one of her top writers. It's soon to launch. Katie is one of the top tech writers. So take it.
Benner:  Thanks, David.
So as David mentioned, we are going to be speaking about platforms, which is a concept that's come up in small ways throughout the conference and in big ways. Mapping, is that a platform yet, gaming, security and research. But we're going to just dig into the concept.
Briefly, I will introduce the panelists. We have Tom Conrad directly to my left, who is the CTO at Pandora. He leads Pandora's product organization, and he began his career at Apple. He holds three U.S. patents, and he still listens to CDs. There is a lot going on with Tom.
Deep Nishar is the senior vice president of user experience at LinkedIn. He's read all of our profiles. He knows everywhere we've ever worked and anybody who ever wanted to hire us. He was previously at Google for six years, and he was a Bakers Scholar at Harvard's business school.
And, finally, last but not least is Charlie Songhurst. Now he is the founder of Katana Capital, which is a longshore hedge fund and the Songhurst Group, which invests in private companies. Previously he was at Microsoft, head of corporate strategy and M&A.
So I'm very excited for this panel. They are all very smart, and they all have lots of idea about platforms.
I want to begin with you, Deep. You had a great definition of what you think a platform is. You said it was two things, we had a conversation earlier. I was wondering if you could unpack that a little bit for the audience.
Nishar:  Do you want to remind me what it was?  After last night's ZZ Ward, I think I'm still—
Benner:  Last night people were by the pool for a long time. I'm not naming names, but I understand. You mentioned that a platform could be defined as a launch pad or as a collaborative tool between companies. I was wondering if you could explain maybe the launch pad idea, like how is a platform a launch pad for other companies.
Nishar:  Yeah. Thanks for the reminder. Forty years ago, I started a company, and it was an enterprise software space, and it was really a platform, because we didn't have an application. And the advice I got from friends and investors as I was making my rounds on Sandhill was, no, no, no, don't call your company a platform company because platforms don't make money so you will not get funded. So we promptly pivoted and call ourselves an application software company.
You fast-forward now, everyone wants to be a platform. I think the definitions of platform are still murky, and I'm suspecting you will have different ones here. The one Katie and I were talking about was in one model, and this is the model that many people think of, the social media platforms, it creates a launch pad for many other smaller companies or products or features to access distribution, and to treat the underlying substrate as a launch pad for what they have offered. So they are taking data and information from the substrate and then they're using the distribution mechanisms to go out. So that's one definition. I think that's an unstable equilibrium, because really the value is asymmetric.
The other more stable definition of platform, something that would survive the test of time, would be the bidirectional value. Both the substrate, as well as the applications that are being created on that platform, there are substantial value by working together symbiotically. I would say that, say, YouTube is a great example of that.
Benner:  Great. Tom, you also spoke about the idea of an open platform versus a controlled platform. I was wondering if you could tell us a little bit more about that.
Conrad:  Yes, I think broadly platforms are as Deep described. When an organization exposes a set of capabilities that a second organization can build upon kind of to the mutual benefit of both entities, you're starting to get to the spot where maybe you can hold your head up and say you have a platform.
There is this separate idea, though, of openness. The Twitter platform, for example, in its initial incarnations was very open. Very few rules about what you could or couldn't do with the platform. By contrast, the Pandora technology platform, there are actually a couple of different kinds of platforms at play inside Pandora. But the technology platform of Pandora, which, for example, enables our partners to deliver Pandora into the context of consumer electronics or automotive integration, I think those fit the definition of platform. We expose capabilities, other people take advantage of them to kind of mutual benefit. But it's closed platform. It's a platform where we carefully choose the partners and vet the implementations that get built on top of our infrastructure. That can work just as well. I think in some instances, be even more powerful than something that's more broadly open. I think Twitter, in fact, has learned that lesson.
Benner:  Why do you think it can be more powerful?
Conrad:  It gets a little bit to this notion of the sustainable model is one where there's mutual benefit. If you have a set of conventions and rules and kind of contracts around how the two parties work together to deliver value, you can kind of ensure that they are mutually beneficial, as opposed to kind of lopsided in one direction or the other.
Benner:  Okay. Charlie, we just mentioned Twitter. Inevitably that was going to come up. When we spoke about platforms, you said the size of platform was important. It has to support a billion dollars, I believe, of other businesses, was the number you came up with. Why is that number so important?  Because I think that would take something like Twitter out of the running for being a platform.
Songhurst:  So I think there's two types of platforms, the aspiring platform and the platform that's made it. The very nature of the internet means that almost every business is an aspiring platform. You think about the difference between now and sort of pre-internet age. Anyone that's got a set of APIs somewhere in their technology that someone else can access is somewhat of a platform. We're all platforms, every single business. So the word sort of loses a lot of its meeting.
I think when you've made it as a platform is when you have a vibrant ecosystem of other businesses that have real valuations and real user bases that depend on your platform. I think a good example is, imagine you switched off your platform and it was inaccessible. What would stop happening. So if you switched off iOS, that would be on the front page of every newspaper in about five seconds. If you switched off Windows, the same. If you switched off YouTube. These are business that have real scale and have huge ecosystems built off them. Lots and lots of business that have achieved decent valuations. I think that is the test. A good test of whether a platform has made it is, are they hosting a conference or something every year where there's entrepreneurs in the audience who have built their business off them. And if not, they're just a wannabe platform.
Benner:  Okay. I want to sort of look at a test case, Google. You have a lot of knowledge about Google, and you said even though the company didn't really think of itself as a platform, it really was. Why is that?
Nishar:  I think Google is moving more in the direction of a deliberate platform. But back in the day, I'm talking the early 2000s, I think Charlie's definition is a great one, which is everyone who has APIs, whether they're open or not, could consider themselves to be a platform. Because theoretically and conceptually, someone else can come and build something on top of that.
I think Google started making that move really with Maps. So Google Search APIs were available for a long time. But at least back in the day when I was still there, they weren't really supported, didn't have a lot of exposure, and people used them—couldn't use them at scale. But Maps was the first place. So Google Maps APIs were available.
Then they started making other sorts of APIs available. YouTube was a very interesting case. It didn't start as a platform. I don't think Steve or Chad, they were building a video dating site for the business history buffs out there and that didn't work out and somehow a Saturday Night Live video went wild, and they are like, hey, maybe that is a good thing for us to do. They have really embraced the power of what a platform could be, to the point where it started out as a place we can share our cat videos to people started getting discovered there, and many music stars have been discovered there to professionals putting their stuff there, and now they have opened up studios everywhere.
So they can become a publishing platform for entertainment, for any of us who want to go create professional quality videos, music videos, entertainment, et cetera, and they are bringing in original programming, as well as other curative programming. That evolution has been really embraced now at Google via YouTube, and more and more of their portfolio assets are becoming platform-like.
Benner:  Now, Charlie, you said you didn't think Google was a platform. I'm wondering why that is.
Songhurst:  So I think if you get a very important app, and obviously search maps is a very important app. You sort of turn into a platform almost by default, because you're going to become sort of systematically important to society that entrepreneurs are going to build businesses off you.
And in some ways, Google truly is a platform. Look at the amount of just SEO businesses built off the platform. It's quite a simple platform. It's not that there's that much functionality enabled. It's not like people are building vertical search engines that take the Google horizontal search and turn that into a vertical business.
I would actually say that YouTube of Google's properties is more of a platform than Google itself, because the search index is an index. It's not like there's a whole bunch of authors in it. Whereas YouTube has a whole bunch of people making a living off submitting videos to YouTube. That, for me, passes the test of a platform in a way that search doesn't.
Nishar:  And Android.
Songhurst:  And Android. Oh, definitely a platform.
Benner:  Taking a step back again. Tom, I was wondering if you could talk a little bit about how Pandora—you've described how it does act as a platform in this closed system. But you had also said that in some ways, it isn't. So I was wondering if you could kind of break down how the company is and isn't and whether or not Pandora will have to choose to become a platform or not as it goes forward?
Conrad:  I spent a lot of time in Silicon Valley with the platform guys. Every company has a few, and I've kind of become part of that community. And it seems that there's a lot of kind of self-worth and identity kind of attached to whether or not you have a platform. I kind of don't care very much, honestly. You know.
We're a pretty pragmatic bunch at Pandora. Our ambition is to define the future of radio. Among the things you have to do to step into those shoes is to be ubiquitously available, you know, in every living room, in every automobile, on every bedside table. So that suggests a certain amount of technology infrastructure that starts to look like a platform that enables people in the automotive domain or consumer electronics domain to build implementations of Pandora that fulfill that objective. Does that make us a platform on the technology side?  I don't know. I'm not really sure I care very much. Our goal is different. Our goal is to make Pandora ubiquitously available.
I think we're kind of increasingly becoming an interesting platform for performing artists. You know, we get hundreds of independent music submissions to Pandora a week that end up played to Pandora's audience where they would never find an audience in traditional radio. We spend lots of times with artists returning data about their music, in the context of Pandora, who is listening to the music, what part of the country are they in, what do they look like demographically. What songs are the most successful region by region. Sets them out to go and tour and talk to promoters and convince them that I really do have 10,000 fans in Albuquerque and so forth.
In a way that is somewhat analogous to the platform capabilities that YouTube exposes to video artists, we're increasingly providing those services to musicians. There's probably a platform dimension there as well. I'm still not sure I really care about whether people view it as a platform or not.
Benner:  And yet the platform guys care.
Conrad:  Exactly. I just want to still get invited to the platform dinner.
Benner:  Are they good?  Are they better than the software dinners?
Conrad:  Sure. I want to go to as many dinners as possible.
Benner:  True. I want to put platform dinners on my list of dinners.
I was wondering, Deep, if you could talk a little bit about how LinkedIn is a platform. Does the company try to define itself that way?  Does it not really care very much, just wants to get out there and serve customers?  What is your take?
Nishar:  Yeah. You know, I've been at LinkedIn about five years, the company is ten years old. So I can't go back in history and say did we start off as a platform. When we started off, clearly Reid, Allen, and the other co-founders had a great view into what LinkedIn would evolve into, which is to really connect talent with opportunity at a massive scale. If you want to do something like that, add large scale, definitionally you are a platform. But let's look at the technical definition. Are there APIs?  Are there things that are happening?
What you find is LinkedIn has become an amazing platform, but in a very different way. Not in a technical sense where there are hundreds of thousands of apps being billed on the LinkedIn platform, although we have with our API program, there are 200,000 plus developers that are using us. But it's not being showcased on LinkedIn. They are using our functionality elsewhere and creating their business back to Charlie's definition, on it.
But the place where we become a real de facto platform is within enterprises. We've become a platform whereby enterprises hire, market and sell whatever they're producing, goods, services, et cetera.
How many people here have recruited or found somebody on LinkedIn through their places of work?  That's quite a few.
You know, it's the de facto hiring platform. It's become a marketplace, bidirectional. Our members who are on LinkedIn get tremendous value because they get all these opportunities. The enterprises who are on LinkedIn are getting tremendous value because they can get passive candidates.
Similarly, LinkedIn with 260-plus million members has an amazing audience for marketeers. So if you want to conduct conversations as a marketeer about your product and services, that's becoming the de facto platform for doing that. So it's become a business platform. We are giving them APIs and all sorts of capabilities to start having those conversations with the audience that they have. But it's a different kind of platform. It's a very powerful platform in order to conduct business, to hire, market and sell within an enterprise.
Benner:  Okay. You know, one of the questions—platforms, the idea, it's a very tech idea. But as technology seeps into more and more industries, whether it's autos, which Tom had mentioned before, or security, suddenly we are seeing tech in other industries. So are we seeing the concept of platform move to other industries. This is something, Charlie, that you had discussed, and I was wondering if you could talk about that a little.
Songhurst:  Sure. First, I'll sort of pick up on Deep's point, which is I think there's a real difference in these two words between marketplaces and technical platforms. I think what we've seen is a huge growth in marketplaces. Before, there were only technical platforms. Think of Windows, think of iOS. Think of an operating system, it's a technical platform. People write code on it.
What you have now is a whole generation of marketplaces, whether it's FMB, whether it's Uber, whether it's LinkedIn, whether it's a dating website. These are all marketplaces of a type. And some of those marketplaces you can make a living off. There's a huge mass of headhunters or recruiting consultants making a living off LinkedIn. Arguably, all of its members make a living off LinkedIn, because it's sort of supplying their resume'.
You think that owning an asset, an apartment that you're renting out, you're making a living off of it?  I think that's conceptually different, though, from building a sort of business. It's different from building, say, the adobe business off the Windows platform back in the '90s or building the Evernote business off iOS. I think that's definitely something worth delving into.
What I would say in the left of the world is you've suddenly got submergence of marketplaces across the world in the variety of industries, particularly around this concept of asset sharing or capacity utilization. Which is you take an asset, and because you know where everyone is because you know the location of their phone and because you know the state of that asset, you can start to share it in a way that hasn't been shared before. So that's obviously the Uber car, that's the ad in the apartment. I had this one with dogs. There's a variety of these businesses. You can see them spreading throughout society.
So there is almost nothing that is in utilization below, say, 20 percent of the time that shouldn't end up in a capacity utilization shared marketplace. I think that's going to be incredibly disruptive to the world. Think of something like the rental car market. How many minutes is a rental car used per week. I think that's going to get disrupted over the next decade. We could go through each industry and see the spread of those platforms. And that's definitely a new type of platform that hasn't existed before.
Benner:  I see Tom and Deep both nodding. Did you want to jump in and add anything?
Conrad:  No, I think that's right. I think teasing apart platforms from marketplaces is a nice way to make the conversation more concrete. Particularly in a world where, you know, there does seem to be maybe an overemphasis in the sort of aspirationally everyone being a platform. I've met plenty of young companies that before they found any consumer traction were hard at work on their API and their quote-unquote platform, and it seemed to me to be an incredible distraction from building that initial set of core values.
The one thing I've been thinking about as we talk about this, is I think we talk about the Windows platform or the Macintosh platform in the '90s and compare that to something like the Facebook platform of the early aughts. There's a pretty fundamental different contract between the provider of the platform and the developers that sit atop it.
I worked at Apple from '91 to '95, and there was this incredible responsibility we felt when we were designing APIs to ensure they were kind of future proof and had room to grow without breaking the participants and the ecosystem. Even when we shipped new versions, we felt an obligation to do, frankly, relatively hackish things to ensure that applications continued to function.
There seems to be no such contract in this new world. I don't know if that's a good or bad thing, but it's certainly interesting. I think it's allowed platforms to evolve more quickly, but oftentimes to the detriment of the people that build their businesses on top of it.
Nishar:  I think it brings up an interesting point. I think the way we think about platforms today is really a semantic construct that's changed over a period of time. And I grew up in India. English is not my first language. So I always think of the meaning of words, and they have a lot of power. I try to be very precise as an engineer in the words I choose. And for the first three years of my life in this country, I would use the word "interesting" because I thought something was really interesting until someone took me aside and said that really means you don't like it, in American English, so don't use the word "interesting."  If you really think something is, use "brilliant" or "great" or "sick" or whatever, right?
Conrad:  Yeah. Every time you think, just go, "Sick!"
Nishar:  So my brain still tries to parse the constructs that we use. I think you brought up an interesting point. Marketplace, they are marketplaces, they are not platforms. But what does it mean?  If you define a construct in a way where there is a mutual contract between the provider and the user, where the user is not just using it all the time but also providing something back, let's say that is a definition, then a lot of things fit into it.
I got my start building engine controllers at a diesel engine company. We were building platforms. We were building energy platforms. Because, by itself, no one buys an engine and does anything with it. We had to work with the truck manufacturers and the chassis and make sure the power trains would fit into the engines to be built. And then the power trains had to fit within the 18-foot tractor trailers that they went into. Until the whole system came together, none of us made any money. That whole thing, we were all building our components of the platform, and we had mutual agreements, we worked closely with each other, just like vendors and OEMs. And that's one of Microsoft's trends is you work with hardware device manufacturers and software providers. And Adobe and Microsoft probably had tons of engineers sitting in each other's offices making sure everything worked well. That contract goes away when a startup starts off by saying, well, I have APIs, so I have a platform, because they don't have the resources to be successful themselves, let alone trying to be a platform where you can create this bidirectionality and this mutual beneficence.
Songhurst:  I think to Deep and Tom's point there is a difference between how much the value to the end user is derived from the platform attributes versus the application attributes. So certain types of platforms, whether it's early Windows, Mac, iOS, Android, force.com, EC2, where they are platforms first and foremost. And the majority of the value derived by the users is from those platform characteristics. And those tend to be responsible platforms, because they are sort of their bread and butter, their food comes from the platform ecosystem. So you tend to have this good, mutually negotiated progress with their providers, and they tend not to sort of eat those providers.
Then you have—and I think this is more in the notes and afterwards—a generation of apps that only when they hit massive scale started to get platform characteristics. And those companies, I don't think feel that same sort of sense of obligation to the platform attributes. If you see a lot of the companies that have built businesses off of them, they've come to somewhat sticky ends, because at some point the platform decided your business is another aspect of my app, it's not a feature, and I'm going to eat your system.
And I think that's an interesting thing for aspiring entrepreneurs, which is if you're depending on a platform, it's good to turn the question back and think how much does that platform depend not on you as yourself, but on you as a class of person.
Benner:  Now we're going to go to questions in a couple of minutes, just so you can start thinking about that. If you would like to ask any one of these panelists something.
Before we do, I just want to touch on this idea of what happens when these platform providers and the companies that build off of them really start to become successful, like Android and Samsung, and they grow and grow and grow, and they create something great that might be even bigger than what they could have done individually. The relationship, though, this collaboration gets a little tricky. And I think we all spoke about this in our interviews a little bit. And I was wondering if anybody wanted to jump in and talk about how that works and what the pitfalls are?  We know what the promise is, sort of domination. But what are the dangers?
Songhurst:  I think the platform providers, you want lots of apps, but you never want one app that's too powerful. I think when one app becomes so powerful, you start to get nervous of it and you start to get an incentive to compete. And that's probably the story of Google and Samsung over the past year. When you start to have most of your distribution coming from one provider, you start to think how can you shift power from them back to you. And that's definitely not a healthy relationship.
Conrad:  It's interesting that the EC2 has no end consumer affordances, right?  The power grid has no end consumer affordances. But the platforms that do, whether it's iOS or Android or Windows, as they look to differentiate their up—in their upgrade cycle, particularly I think when the upgrade cycles were paid. You know, you have a revenue number to hit by shipping a new version of the Mac OS or whatever. There's this temptation to take the most successful interesting applications that get built on your platform and incorporate them into the operating system. That's what happened with internet Explorer, of course.
And so that's another characteristic to look for, I think, if you're building on top of someone else's platform is, you know, think about their business model, think about how they differentiate their product over time and how likely it is they might move into the space that you hope to enable on top of their capabilities.
Nishar:  To paraphrase Henry Kissinger in this context, there are no friends or enemies in this business, there are only permanent business interests.
Conrad:  You think about the Twitter case with respect to their decision to take over the development of mobile clients. And as—if you were a developer of a mobile Twitter client, sitting on top of the platform, it's hard for me to understand how you would really convince yourself that Twitter wouldn't ultimately move into that space.
Nishar:  You could hope they buy you out.
Conrad:  Yeah, right. But short of that possibility, it's hard to conceive of a world where you view, like, five years down the road, ten years down the road Twitter will have continued to decide that the best way to expose their capabilities to end users is to do it through a myriad of apps that don't use any consistent experience, don't monetize. I mean, I guess you could go to bed every night telling yourself that Twitter is like the power grid. That the future of Twitter is really about delivering bits, and they don't care about the clients that sit on top of it. But it seems a little, I don't know, delusional.
Benner:  I thought I saw a hand go up over here. Yes.
Lewin:  Yeah, Dan'l Lewin with Microsoft. I really like the way you're framing the conversation, but I would ask a question that I think you're missing. Maybe I think you are. There's a difference between content less and content rich, so the question of whether the content is present upon which other people do things or whether the applications or the relationship between the platform and the partner is content less, right?  So there is user generated that then can be—who owns the data?  That's sort of the core question that I'm sort of getting at. Do you understand the core of my—I don't know that I'm framing it well. If you think about Pandora, there is content that you present that's already been juried, that has copyright. And then there's those that want to openly present it. LinkedIn people are putting the content in, Facebook people are putting the content in. Same with Twitter.
But the other platforms are content less, and then most of the applications that got built on the early ecosystem were also content less, right, automating rational tasks. So maybe can you chat about that a little bit?
Nishar:  I think I would start with Dan'l—see if it addresses your question. I think content is certainly an important part, and that is something that—when people build applications or any platform where there's an exchange of value, content can be a big part of it. What we find is that our platform at LinkedIn, the content that people really want is identity. The reason they want identity is multiple, and we only give it with the user's explicit permission, is many times they want to reduce the friction of having a user on board their particular app. So logging with LinkedIn is one of the easiest ways to do it, especially when it's a professional application. And so the context is just right.
In return, like we have maybe 1-plus million share buttons across the Web on content domain sites. It's like curated third-party content. But those sites, when they have the share button which enables any user to share frictionlessly, which then brings them virality and more users on their site, we get rich, interesting content in a metadata sense, which tells us that this is what people care about in this particular industry or this network, and then we can present that content back to their networks and to other users on LinkedIn. So it becomes a good symbiotic relationship from that standpoint. So there is some content changing hands, but the format of that is very different. We are sending an identity-related information, and they are sending us back actual content, which then generates more value for them and it generates value for our network, because our network is getting rich, professional content, which enables them to be better at what they do every day.
Benner:  I think we have time for a couple more questions. Over here.
Bonchek:  Hi, Mark Bonchek with ORBIT. My question is following on the discussion about how the platform idea might be extending into other industries, and if the test is that people can build their business on it, then there are plenty of solo entrepreneurs who in effect use Starbucks as their office. So does that mean that Starbucks is a platform?
Benner:  Who would like to take that one?
Conrad:  In the sense that this stage is a platform.
Nandlall:  The stage is called a platform. It's a construct.
Conrad:  No, it is literally a platform.
Benner:  Another question?
Nishar:  Yeah, I think if you go further, Starbucks could go away, use Charlie's definition. But some other coffee shop may have all those people show up. So maybe it's a platform, but in a femoral sense. You can live without it. So maybe it's not an enduring platform, from that standpoint.
Bonchek:  They put WiFi in, so there was a technology layer.
Nishar:  Yeah.
Conrad:  The Starbucks CIO can come to the platform dinner if he wants.
Benner:  Then we'd all want to go to the platform dinner.
Songhurst:  After the platform dinner, it does bring up a great point, which is, is this something too many people aspire to. Like, should we say that as an entrepreneur, 99 percent of the time you should ban the word "platform" until you've hit a billion dollar valuation or something. Because is it an easy delusion to fall into when you start building platform features, when you should be building end-user features?
Conrad:  I personally think any time you expand the scope of your company to embrace a new class of consumer, and you need to provide products for them, you need to support them, you need to fulfill your contract with them, whether it's explicit or implicit. Yeah, that's a defocusing moment. You can only afford to do that so many times before you run the risk of completely losing your way.
Nishar:  Yes. And if everyone is going to be a platform, who is going to be the application?  Who's going to serve us?
There can be benefit. Like think about it. The guy in the brown uniform who delivered packages at our doorsteps was just the UPS guy, and now he's a global logistics support specialist. Right?  He is. But the fact of the matter is that UPS, that's not just a phrase or branding campaign. They actually literally are going into large organizations saying we can take care of your logistics issues. And in the days of Amazon and the vast majority—not the vast majority, but a material portion of our commerce going online, that is a very interesting evolution for UPS. And that is the right evolution for them, I would contend, to become a true logistics platform.
Benner:  Sort of a rapid-fire question just quickly, I'm going to ask each of you to name me what—a company that you think is currently sort of the most successful big platform and then a company that is emerging as a platform that you think is interesting to watch. I don't know who wants to start off. But you all have to answer.
Songhurst:  I will start with the emerging one. I think Bitcoin in digital currencies is a fascinating area that may or may not be a platform, but it certainly has the characteristics in the sense of the concept of digital currency is something that obviously other businesses will build off. It's globally scalable, which I think is one of the key things to make it a successful platform. So I think that definitely it could be the space to watch. Emerging.
Benner:  And then an established?
Songhurst:  Dan'l is here, so I will pick Microsoft as the established platform.
Benner:  No bias there at all.
Nishar:  I would say established is YouTube, and emerging is Vchat.
Conrad:  Just to make it fair, I will go with iOS as established. And I think over the course of the next ten years, all of our cars are going to be connected to the internet in the same way that the devices we carry around in our pockets are connected to the internet, and the automotive industry is going to have to solve the platform question for that environment to come into its own. So I think really interesting to watch. Lots and lots of companies working really, really hard in that space today.
Nishar:  Could I actually add one more to the emerging one?
Benner:  Yeah.
Nishar:  I was at the media lab earlier this year, and I saw a really interesting project called LuminAR. I don't know how many people here know about it. But effectively, every electric socket in the world can be its own API. If you have the same voltage and the same amperage, you can attach any device to it. There are probably a couple billion of those in the U.S. alone. Can you imagine if you started integrating with that?  You can do all sorts of very interesting things. And going back to the electric grid as the platform, I think the LuminAR project really converts every electric socket in this country and beyond as an amazing platform that could do a lot of interesting things for us if the NSA doesn't get there first.
Benner:  This is fantastic. We are out of time. Thank you so much for joining us, and I hope you have a great rest of the day.